Wednesday, January 11, 2012

Atrial fibrillation

Author : Dr Gregory Marcus Cardiac Electrophysiologist University of California, San Francisco

2008-08-01

Normal Heart Function
In order to understand atrial fibrillation, it is first important to become familiar with the basic anatomy and physiology (or function) of the heart. To review, the heart has 4 chambers: the top 2 chambers, called the atria, are relatively small and thin; the bottom 2 chambers, the ventricles, are larger and thicker. Blood that has drained all of the body’s tissues (except for the lungs) flows through the veins into the right atrium. The right atrium then contracts, propelling blood across the tricuspid valve and into the right ventricle. The right ventricle then squeezes that blood across the pulmonary valve, through the pulmonary artery and into the lungs. Once in the lungs, the blood is re-supplied with oxygen and continues on its way via the pulmonary veins into the left atrium. The left atrium then contracts, pushing the blood across the mitral valve and into the left ventricle. Finally, the left ventricle squeezes the blood across the aortic valve, through a large artery called the aorta and on to perfuse the tissues of the body. 
The normal electrical conduction system of the heart
           The right and the left sides of the heart actually work simultaneously, so that the atria both contract together, and then the ventricles contract together. These coordinated pumping actions are orchestrated by the heart’s electrical conduction system, and the sinus node can be thought of as the conductor of that orchestra. The sinus node is the normal or physiologic pacemaker. From this structure, situated in the upper right atrium, the electrical signal responsible for each heart beat arises. Normally, in response to signals from the brain or the adrenal glands, the sinus node will speed up as needed (such as during exercise, anxiety, or excitement) and slow down when appropriate (such as during rest or sleep).
            How does that sinus impulse activate the heart? It is important to understand that the muscle cells of the heart (the same cells responsible for contracting and relaxing the chambers) can conduct electricity from one cell to the next. For example, an electrical signal that starts in one part of the right atria can travel from cell to cell throughout the right atrium and into the left atrium (the right and left atrium are immediately connected and share the same septum). Similarly, an electrical impulse that starts in one ventricle can propagate from one ventricle to the other. Normally, the atria are separated from the ventricles by a fibrous structure which houses the tricuspid and mitral valves, making the 2 atria electrically disconnected from the 2 ventricles. In fact, in a normal heart, there is only one way for an electrical signal in the atria to communicate down to the ventricles, and that is via the AV node. The AV node is situated in the middle of the heart, receiving electrical connections from the atria and delivering electrical connections to the ventricles.
            When a heart muscle cell is activated by an electrical signal, it contracts. In sum, a normal heart beat arises in the sinus node and propagates across the right atrium and the left atrium via cell to cell electrical connections, resulting in contraction of both atria nearly at the same time. When it reaches the lower atrial septum, it then continues to the AV node. After traveling through the AV node, this electrical signal continues down a network of specialized conduction tissue that spreads the impulse in an organized fashion throughout the ventricles, resulting in an organized contraction of each ventricle nearly simultaneously (Figure 1).
 Figure 1. The normal conduction system of the heart. Yellow arrows demonstrate an electrical signal originating in the sinus node and propagating through both the right and left atria. This signal will reach the AV node and travel through the ventricular conduction system (shown in gold), ultimately electrically activating the right and left ventricles.  The left atrial appendage, a potential site of blood clot formation in atrial fibrillation, is also shown. This figure was obtained with permission from Mr. David Criley at www.blaufuss.org.

What is Atrial Fibrillation?        

As might be obvious from the name, atrial fibrillation is present when the atria are fibrillating (Figure 2). This means that the atria are conducting electrical signals in a disorganized, erratic, and rapid way. Therefore, as the muscular contraction of the heart follows the electrical activity, the atria are contracting in an equally disorganized way. In fact, different parts of the atria (left and right) are contracting and relaxing in a completely uncoordinated manner. Although the exact electrical phenomenon responsible for this rapid and irregular conduction is not yet completely understood, there are 2 general explanations1
: one is that there is a focus somewhere in the atrium that is firing very rapidly. Because the atrial tissue is not all able to conduct at such a rapid rate, the impulses reach different parts of the atrium at different times, resulting in a rapid, but disorganized and irregular beating of the atrium. The second explanation is that multiple partially reentrant, or circularly traveling, “wavelets” of electrical excitability are present, propagating around the atrium in a random fashion. It may be that some patients have more of the rapid focus as a cause and others have more of the multiple wavelets as the cause. It is also likely that many (if not all patients) have both- for example, the rapid focus may trigger the atrium to ultimately form multiple electrical wavelets.
Figure 2. A heart in atrial fibrillation. The yellow arrows illustrate the chaotic and rapid nature of atrial activation. The gold circle in the upper left atrium represents a possible focal source of the arrhythmia (near the left upper pulmonary vein). Note that activation of the ventricles continues to occur via the AV node. This figure was obtained with permission from Mr. David Criley at www.blaufuss.org.
             In general, there are 4 direct consequences of atrial fibrillation:
·      Loss of the atrial contraction
·      A rapid pulse
·      An irregular pulse
·      A risk for stroke
Loss of the atrial contraction: Because the atria are no longer electrically activated in an organized fashion, the atrial contribution to ventricular filling (sometimes called the atrial kick) is lost. Flow does continues to occur from the veins into the ventricles (driven primarily by a pressure difference) and so this condition is by no means immediately life threatening. However, because the efficiency of ventricular filling is diminished, a patient with atrial fibrillation may experience fatigue or shortness of breath.  In some conditions where the ventricles need as much help as they can get with filling (such as when someone has a ventricle that is stiff due to a previous heart attack or long-standing high blood pressure), the patient may become very symptomatic with loss of the atrial kick. At times, this can be severe enough to result in an increase in pressure in the veins filling the heart, ultimately leading to fluid build-up in the lungs and other tissues of the body, resulting in something called congestive heart failure. Of note, if the ventricles are very compliant (ie, easy to fill), some patients may not experience any problems due to the loss of the atrial kick.
A rapid pulse: At rest, the sinus node normally generates an electrical signal approximately 60-100 beats per minute. As the pulse is determined by ventricular contraction (actually by the left ventricle, which is pumping blood to the body) and as the sinus impulse is normally propagated down through the AV node and into the ventricles, the normal resting pulse is also then approximately 60-100 beats per minute (with normal heart rates often even slower than this). When a catheter that can record electrical activity is placed inside an atrium during atrial fibrillation, it will record irregular electrical activity that can vary considerably in rate, but is most often faster than 350 signals per minute. Therefore, in atrial fibrillation, the AV node is being bombarded with these rapid impulses at rates greater than 350 times a minute. Part of the normal function of the AV node is to protect the ventricles from such a fast rate, and it therefore does not let every one of those rapid electrical signals through. However, it is not always perfect, and so the majority of the time it allows enough of those signals through to result in a rapid pulse, often between 100-200 beats per minute. This rapid pulse is likely responsible for the majority of symptoms a patient feels when in atrial fibrillation: fatigue, shortness of breath, and palpitations. In rarer circumstances, chest pain and feeling faint can be caused by an inappropriately fast heart rate.
An irregular pulse: Although the AV node does not propagate everyone one of those rapid impulses arising from the atria during atrial fibrillation, it is still being activated in an irregularly irregular tempo (as opposed to regularly irregular, which would at least have some kind of pattern to it).  Therefore, in atrial fibrillation, the activation of the ventricles, and therefore the pulse, is irregularly irregular in timing. It turns out that this irregular rhythm itself also adversely affects the efficiency of the heart.2 Although probably not as important as loss of the atrial kick and the rapid pulse, the irregular pulse itself may contribute to a patient’s symptoms. In addition, it is often this irregularly irregular pulse (either palpated by a physician or heard when listening to the heart) that signals the possibility of underlying atrial fibrillation. It is important to note that, although atrial fibrillation is probably the most common cause of an irregularly irregular pulse, other heart rhythm problems (some more benign, some more worrisome) can result in a similar kind of pulse.
A risk for stroke: In the general sense of the term, a stroke occurs when blood flow is impeded to a section of the brain for a sufficient amount of time so as to result in loss of function. For example, if the part of the brain responsible for moving the left arm loses its blood flow for a certain amount of time, paralysis in the left arm will result. One of the most common causes of stroke is thromboembolism: “thrombo” is related to the word “thrombus,” or blood clot. “Embolism” refers to something in the blood stream that moves from one area to another. Hence, thromboembolism occurs when a blood clot moves from one place in the body to the other via the blood stream.  
            Whenever blood ceases to flow, it tends to clot. In the fibrillating atria, when the normally vigorous atrial contraction is lost, blood is no longer flowing as it should and instead can become static. Therefore, a blood clot is more likely to form in the fibrillating atria. If such a blood clot forms and then dislodges into the circulation, it can reach the brain and result in a stroke. It is well known that people with atrial fibrillation are at a significantly higher risk for stroke.3 It is important to understand that the risk of stroke in an atrial fibrillation patient varies considerably and is predominately determined by the presence or absence of other stroke risk factors (discussed below in section VI.2.). 

Who gets Atrial Fibrillation?
           
Atrial fibrillation is the most common abnormal heart rhythm (or arrhythmia), affecting several million Americans.4 While essentially anyone can get atrial fibrillation, there are known established risk factors for the disease. The most important risk factor is simply age: of everyone over age 60 in the United States, approximately 4% have atrial fibrillation; of everyone over age 80, approximately 10% have atrial fibrillation. Other important risk factors include a history of high blood pressure (or hypertension), heart failure (meaning the ventricles, particularly the left ventricle, either does not contract strongly or is too stiff to fill properly), and heart valve diseases. In addition, men are at higher risk than women and whites are probably at higher risk than other races. People with coronary artery disease or a previous history of a heart attack also appear to be at higher risk. Finally, people with lung disease and diabetes can also be at higher risk.

Are there different types of atrial fibrillation? 
            There are indeed different types of atrial fibrillation, and the disease can be categorized based on the behavior of the arrhythmia or based on the characteristics of the patient.1
            Atrial fibrillation can be paroxysmal (in other words, it comes and goes on its own), it can be persistent (meaning that a person is in it all the time and would require some kind of treatment by a physician to get out of it), and it can be permanent (meaning that, even with available therapies, the atrial fibrillation will not go away). Paroxysmal atrial fibrillation can be quite unpredictable and the timing, frequency, and duration of episodes can vary substantially from person to person: some may have one episode that lasts 10 minutes once a year, while others may have multiple episodes a week, lasting for hours at a time. Some people can identify triggers for the onset of their atrial fibrillation, such as alcohol, exercise, or sleep- these also can vary substantially from patient to patient. A paroxysmal atrial fibrillation patient can sometimes develop persistent atrial fibrillation and, after a persistent episode is converted to a normal rhythm, paroxysmal episodes may develop. Those with permanent atrial fibrillation typically have had persistent atrial fibrillation for a long time as, in general, the longer a person is in atrial fibrillation, the more difficult it is to get them out of it.
            Based on characteristics of the person, atrial fibrillation can be “lone,” meaning that the person has none of the risk factors listed above: in general, if someone is younger than 65 and has no known heart or lung disease (including hypertension or diabetes), they have lone atrial fibrillation. In general, as discussed below, these people are felt to have a much lower risk for stroke and probably have a generally better prognosis than those with other established heart or lung risk factors.5
            It is also well known that atrial fibrillation can commonly occur after surgery, particularly after there has been any kind of surgery on the heart such as coronary artery bypass surgery or a valve replacement. This post-operative atrial fibrillation is typically felt to be self-limiting and may not reflect a long-term risk for the disease. Of note, those with lone atrial fibrillation or post-operative atrial fibrillation can have paroxysmal, persistent, and even permanent atrial fibrillation.
            Finally, some individuals are thought to have vagal atrial fibrillation.1 To understand this, first one must be familiar with the autonomic nervous system, the part of the nervous system that tends to bodily functions that are not directly under our conscious control, such as our heart rate, blood pressure, digestion, sweating, etc. The autonomic nervous system is comprised of and constantly influenced by a balance between the sympathetic and parasympathetic nervous systems. The sympathetic nervous system involves adrenalin, or the “fight or flight” responses, such as a faster heart rate, a higher blood pressure, excitement, and anxiety. The parasympathetic nervous system opposes the sympathetic and generally is more active in the resting state, lowering the pulse and blood pressure, facilitating gut motility and digestion. The primary nerve involved in the parasympathetic nervous system is the vagus nerve; therefore, parasympathetic activity is also often referred to as ”vagal.” While some individuals appear to have their atrial fibrillation triggered by exercise or sympathetic nervous system activity, it seems that even more patients reliably have a vagal trigger for their atrial fibrillation. These are the people that experience atrial fibrillation while asleep, while relaxed, or often after exercise. In some, atrial fibrillation can actually be terminated with exercise. Often, these atrial fibrillation patients are very fit and otherwise quite healthy. It may be that an exaggeration of vagal tone (known to occur at rest in healthy individuals and otherwise felt to be a very healthy condition) makes individuals more prone to atrial fibrillation.

How is the diagnosis of atrial fibrillation made?

            The majority of the time, the diagnosis is made by an electrocardiogram (ECG). The electrocardiogram involves having several electrodes placed on the body. It is painless and typically takes less than 5 minutes. With this test, the electrical activity of the heart can be seen: a small wave, called the P wave, corresponds to electrical activation of the atrium during normal sinus rhythm (Figure 3). In a normal rhythm, the larger QRS complex, representing electrical activation of the ventricles, follows shortly after the P wave is inscribed. In atrial fibrillation, because there is no organized atria activity (as described in sections I and II above), there are no P waves. In fact, the fibrillatory action of the atria are instead seen as a constantly undulating line, with no regular pattern (Figure 3).  The timing of the QRS complexes are irregularly irregular, corresponding to the timing of ventricular activation (and therefore the pulse) described above.


Figure 3. The top recording demonstrates normal sinus rhythm, with deflections called P waves (denoted by asterisks) that represent normally conducting atria. Each P wave is followed by a QRS complex, representing ventricular depolarization (solid arrows). Each QRS complex is followed by a T wave, representing repolarization of the ventricles (dashed arrows). The bottom recording is from the same patient while in atrial fibrillation. Note the absence of P waves and a subtle irregular undulation of the baseline representing atrial fibrillation. The QRS complexes (ventricular activation) occur more rapidly and in an irregularly irregular pattern. 

What are the consequences of atrial fibrillation?      


There are generally 3 consequences of atrial fibrillation that available therapies address: symptoms, a prolonged rapid rate, and a risk for stroke.
            As described above (Section II), a combination of loss of the atrial kick, an irregular rate, and an inappropriately rapid pulse can all contribute to symptoms. Symptoms can vary from patient to patient, but typically include fatigue, decreased exercise tolerance, palpitations, and/ or shortness of breath. Not all patients are symptomatic, and some are clearly much more symptomatic than others. Apart from preventing long-term damage to the heart or the body, just the subjective improvement or elimination of symptoms is an important goal in the treatment of atrial fibrillation (as below).
            In addition to contributing to a patient’s symptoms, a prolonged rapid ventricular rate can itself damage the heart. If a patient’s heart rate (meaning the pulse or ventricular rate) is persistently great than 110-120 beats per minute over months on end, a tachycardia induced cardiomyopathy can develop. This means that the ventricles can actually become dilated and weak, leading to congestive heart failure (where fluid fills up in the lungs and the tissues of the body) and an increased risk for other potentially life threatening arrhythmias in the ventricles. Therefore, independent of symptoms, it is important to maintain rate control of the ventricles, preventing long-term elevation of ventricular rates.
            Finally, as described in Section II, status of blood flow through the fibrillating atria can lead to the formation of a blood clot which, if dislodged, can travel and block a blood vessel supplying the brain, leading to a stroke (Figure 4). Independent of treating symptoms and achieving rate control, prevention of stroke is always an important consideration in the treatment of atrial fibrillation. 
           
Figure 4. This represents a computed tomographic scan (CT scan) of a patient with atrial fibrillation and a stroke due to embolization of a clot from the left atrium. Panel A shows the 3-dimensional reconstruction. Importantly, the primary vessels supplying the right side of the brain appear normal (white asterisk). Panel B is a cross section of the brain (as if looking straight up at a horizontally cut “slice”), with contrast dye injected to light up the blood vessels supplying the brain in white. Note the major blood vessel on the right (the patient’s left) as denoted by the red arrow. As with the other blood vessels seen, a symmetric counterpart to this vessel should normally be observed on the other side. However, as denoted by the large white arrow, no such artery is seen, demonstrating that this vessel (the right middle cerebral artery) has been occluded, resulting in loss of blood flow to the part of the brain that vessel supplies. Panel C is a cross section of the heart (again, as if looking straight up at a horizontally cut slice) at the level denoted by the dotted line in panel A. Because of the contrast dye that has been injected, the inside of the heart and blood vessels should appear grey or white. Note the black area denoted by the white arrow- this is in the middle of a part of the left atrium called the left atrial appendage and represents a blood clot. This image is provided courtesy of Dr. Max Wintermark, Neuroradiology Section, University of California, San Francisco.

Treatment of Atrial Fibrillation


1. Rate versus rhythm control            

One way to think about the treatment of atrial fibrillation is to split it into 2 general treatment strategies: rate control, where one allows the patient to remain in atrial fibrillation, but concentrates on making sure the rate is well controlled, versus rhythm control, where the strategy is aimed at achieving and maintaining normal sinus rhythm.
            Given everything written above, it would seem intuitive that a rhythm control strategy would make the most sense. However, large randomized studies where patients assigned to rate control were compared with those assigned to rhythm control, failed to demonstrate any benefit of one strategy versus the other.6, 7 How can this be, given all of the adverse effects of atrial fibrillation that are so well known? There are several explanations: first, perhaps most importantly, the means we have to achieve and maintain sinus rhythm are suboptimal. Both poor effectiveness and toxicities limit the use of many of the available therapies (see below, Section VI.b.i.-iii.). In fact, one analysis suggested that, if normal sinus rhythm was actually achieved, outcomes were in fact improved.8 The point is that there is a big difference between undertaking a strategy of trying to achieve and maintain sinus rhythm and actually being able to do it. Of note, none of these large studies assessed ablation therapy (discussed below, Section, VI.b.iii.) as a means of maintaining normal sinus rhythm, and such studies are currently ongoing.
            Second, patients with significant symptoms were generally not enrolled in the studies. Therefore, although not clearly proven in large trials, clinical wisdom and experience dictate that severe symptoms of atrial fibrillation (that is, symptoms that persist despite adequate rate control) warrant a rhythm control strategy.
            Third, subjects in the rhythm control arm were often taken off warfarin, the best drug for stroke prevention, when it appeared that they were indeed maintaining a normal rhythm during follow-up. However, the stroke rate was the same in the rhythm and rate control groups.7 This finding has been attributed to the now well documented fact that patients can have asymptomatic episodes of atrial fibrillation. In fact, even in those who are clearly symptomatic during some episodes, a wearable event monitor (essentially, a portable electrocardiogram) demonstrates episodes of atrial fibrillation without the patient being aware.9 Therefore, it is thought that many of these patients who failed to demonstrate evidence of atrial fibrillation either by their own report of symptoms or during their doctor visits likely had asymptomatic episodes that were not documented. Given this data, a rhythm control strategy is likely not sufficient to negate the need for warfarin (or anti-stroke) therapy. However, if one performs rigorous monitoring (such as with a wearable or even an implantable event monitor that captures every heart beat over a month or longer), some believe withholding warfarin in those with no evidence of atrial fibrillation might be safe. For now, this is a decision to be made based on clinical judgment, and future studies may help guide us in determining if such monitoring can indeed identify those at a sufficiently low risk for stroke.

a. Rate Control
While it is useful to divide general treatment strategies into rhythm control versus rate control, in reality the great majority of atrial fibrillation patients undergo some form of rate control. In some cases, the decision is made that a rhythm control strategy is either unsafe or so unlikely to work that a rate control strategy alone makes the most sense. In others who are undergoing rhythm control, intermittent episodes of atrial fibrillation may yet still occur, and it is important to have the ventricular rate adequately controlled during those episodes. Finally, in the majority of patients with atrial fibrillation presenting to the emergency department or admitted to the hospital, the first line of therapy is typically rate control. In fact, many of these patients will spontaneously revert to sinus rhythm within the first 24 hours of their presentation and it is primarily a matter of reducing their ventricular rate until that happens.
            To review some of the discussion above (Section V.), we control the rate primarily for 2 reasons: to control or eliminate symptoms (of palpitations, fatigue, or shortness of breath) and to prevent damage to the heart (prolonged elevated rates can result in a weakening of the heart). In patients with underlying heart disease, rate control can be crucial. For example, in a patient with blocked coronary arteries (or coronary artery disease), an inappropriately increased ventricular rate will increase oxygen demand while potentially decreasing supply (at faster rates, the coronary arteries have less time to fill per heart beat); in a person who just barley receives enough blood supply (and the oxygen the blood carries) at a normal heart rate, atrial fibrillation with a fast ventricular rate can result in an oxygen demand that exceeds oxygen supply, resulting in the death of heart cells, also called a myocardial infarction or heart attack. Similarly, in a person with heart failure, either because the ventricles are weakened in their pumping ability or too stiff to fill adequately, a fast ventricular rate during atrial fibrillation can exacerbate the situation, leading to an acute episode of congestive heart failure, with lungs and body tissue filling with fluid.
            Rarely, a person may not require rate control. Most commonly, this is seen in elderly people who have some degree of natural slowing in their heart’s conduction system due to fibrosis (essentially a build up of scar tissue). In fact, this scarring of the conduction system that occurs with age is typically unrelated to other heart disease and is the most common indication for a pacemaker. Therefore, if the AV node and conduction fibers traveling from the atria to the ventricles (described above, Section I.) are unable to conduct an impulse faster than 70 times a minute for example, atrial fibrillation can never result in a fast ventricular rate. The other circumstance is even more rare: some individuals that experience vagal atrial fibrillation (described in Section III) are under a strong influence of the anti-adrenalin autonomic nervous system during their episodes, an influence that can slow AV nodal conduction sufficiently so as to prevent a fast ventricular response.
i. Medicines for rate control
Rate control is aimed primarily at slowing the conduction of the AV node. Therefore, while the atria remain fibrillating at the same rate, medicines that slow conduction through the AV node can effectively prevent a fast ventricular rate. Because they only slow conduction, they do nothing to regularize the rhythm, and the ventricular response remains irregularly irregular. The agents that are most effective are beta-blockers and calcium channel blockers. Both of these agents lower blood pressure and in fact are used in many patients solely for that purpose. Examples of beta-blockers (generic names) include atenolol, metoprolol, and carvedilol. Different beta-blockers vary in the exact receptors blocked, absorption in different tissues, and metabolism. They slow both the SA node and AV node primarily by blocking sympathetic nervous system influences. They have proven to be very safe and, in fact, save lives by preventing recurrent heart attacks and sudden death in patients who have had a previous heart attack. In addition, they have also been found to prevent death and help the heart to heal in certain patients with heart failure. Beta-blockers are generally well tolerated, but have been reported to cause fatigue, erectile dysfunction, and depression.
            Only certain calcium channel blockers, namely non-dihydropyridine calcium channel blockers, are effective in slowing the AV node and therefore the ventricular rate during atrial fibrillation. Verapamil and diltiazem are 2 examples, both slowing AV nodal conduction via a direct effect on calcium channels. By blocking calcium channels in the heart itself, both can also potentially worsen the pumping function of a ventricle that is weakened at baseline. By blocking calcium channels in the gut, both agents can exacerbate constipation.
            In general, these agents are both quite effective and tend to be tolerated without any difficulty. Because both lower the blood pressure, hypotension (a blood pressure that is inappropriately low) can sometimes limit the dose that can be given. In some rare patients with very low blood pressure at baseline, the safe use of these agents may be precluded.
            An additional agent that can be used for AV nodal blockade is digoxin. This medicine is not recommended as a first line agent because it tends to be less effective than either beta-blockers or calcium channel blockers.1 However, as it does not lower blood pressure or exacerbate poor pump function of the ventricles, it may be considered as first line in some patients. However, toxicity due to overdose, kidney failure, or interactions with other drugs can result in potentially lethal abnormal heart rhythms. 
            Some other agents that are typically used to prevent atrial fibrillation and maintain sinus rhythm (discussed in more detail below, this Section VI.1.b.ii.) also slow AV nodal conduction. One of these is sotalol, which has beta blocking activity. The other is amiodarone, a very complex and potent antiarrhythmic agent. Amiodarone is somewhat unique in that, like digoxin, it can block the AV node without significant effects on lowering blood pressure.
Tachy-brady syndrome
            Not surprisingly, a side effect of all these medicines is that the pulse can become too slow. This is of particular significance when a person has intermittent (or paroxysmal) atrial fibrillation; in that circumstance, they may go very fast during atrial fibrillation, requiring high doses of AV nodal blockers (such as calcium channel blockers or beta-blockers) to prevent the pulse from going to fast, but, when in sinus rhythm, those same agents (which also slow conduction of the sinus node) can result in too slow of a heart rate. In fact, in many patients, the same scarring of the conduction system that occurs with age may affect the atria and make the same person more prone to atrial fibrillation. Therefore, it is not uncommon to see a person with very slow heart rates during sinus rhythm (typically attributed to age-related scarring in or around the sinus node) with intermittent atrial fibrillation (perhaps due to scarring of the atrium). Often in these people, termed as having sick sinus syndrome, the AV node is healthy, resulting in a rapid ventricular response during atrial fibrillation. Such patients are also often deemed as having tachy-brady syndrome: “tachycardia” means a fast heart rate and “bradycardia” means a slow heart rate, and these patients exhibit both. Treatment can often be quite challenging in this circumstance as the medicines necessary to prevent fast heart rates during atrial fibrillation can exacerbate the bradycardia that occurs with sinus rhythm. When this is an issue, the treatment is to place a permanent pacemaker, which prevents the heart from going to slow.10 Then, medicines can be given to prevent the heart from going to fast, and the patient remains protected against bradycardia.
ii. AV nodal Ablation
Sometimes, a patient either cannot tolerate AV nodal blocking medicines or such medicines are ineffective at preventing fast heart rates during atrial fibrillation. Reasons for not tolerating AV nodal blockers are mentioned above: briefly, this would be due to too low of a blood pressure on the medicines or too slow of a heart rate during episodes of sinus rhythm. When either AV nodal blockers are not tolerated or are ineffective, there is a straightforward and very effective procedure that can be performed: AV nodal ablation. Because this results in permanent heart block (ie, no electrical connection between the atria and the ventricles), a pacemaker must always be placed. This procedure is performed under moderate sedation, meaning that an intravenous line (an IV) is placed and medicines to make a person comfortable and usually a bit sleepy (but not unconscious) are given. Some numbing medicine (such as lidocaine or xylocaine) is introduced into the skin in the groin (typically the right groin) overlying a major vein that runs from the leg into one of the big veins that empties into the heart. Using landmarks of the external body, the operating physician then introduces a needle into that leg vein (called the femoral vein). Leaving the needle there, a floppy wire is then threaded through the needle and into the vein. The needle is taken out, leaving the wire in place, and a long plastic tube, termed a sheath, is place over the wire. The wire is then removed, leaving the sheath in the vein. The sheath has a one-way valve in it, such that blood can not flow from the vein outside, but other long plastic catheters can be introduced into it from outside. After all of the air has been removed from the sheath by drawing back blood, removing any air, and flushing with fluid, an ablation catheter is introduced through the sheath, up the vein, and into the heart under visualization by an X-ray camera. This ablation catheter has 3 capabilities: 1) It can record electrical activity, 2) It can pace the heart, and 3) It can deliver radiofrequency energy, which can burn and kill heart tissue. Under X-ray guidance and by interpreting the electrical signals recorded by the catheter, the area of the AV node can be identified. Once there, radiofrequency energy is delivered until the AV node can no longer conduct electrical signals from the atria to the ventricles (Figure 5). This procedure is highly effective and will prevent atrial fibrillation from ever resulting in a rapid ventricular response or a rapid pulse. Studies have shown that this procedure can result in an improvement in quality of life and exercise tolerance for some patients with atrial fibrillation.11 After this procedure, medicines are no longer needed to slow the heart rate.

Figure 5. This is a radiograph (or X-ray image) taken during an AV node ablation procedure. The wire arising from above (dashed arrow) is a previously placed pacemaker lead that enters from a vein in the arm, continues down the large vein that enters the top of the right atrium (the superior vena cava), continues through the right atrium and then is screwed into the high interventricular septum in the right ventricle. The electrode catheter entering from below (solid arrow) arises from a vein in the leg, through the major vein that empties into the bottom right atrium (the inferior vena cava), and continues to be placed over the area of the AV node. The exact location of the AV node is found by analyzing signals recorded from the tip of this catheter. The tip of this catheter is also capable of delivering radiofrequency energy, thereby burning/ablating the AV node and ceasing any electrical conduction through it. (This image is provided courtesy of Dr. Nitish Badhwar, Cardiac Electrophysiology Section, University of California, San Francisco.)
 
            It is important to understand 2 things regarding this procedure: Once done, the patient is absolutely dependent on the pacemaker. Typically, the procedure is done with the goal of leaving the patient with an “escape” rhythm, or an automatic focus that will generate a ventricular beat should the pacemaker fail, but this can not be guaranteed. Second, this procedure does nothing to eliminate the atrial fibrillation itself and so the patient’s atria continue to fibrillate as they would have prior to the procedure. Of note, because the patient is dependent on the pacemaker and because all electrical communication is interrupted between the atria and the ventricles, the ventricular rhythm (and therefore the pulse) is actually regular. As noted above (Section II.), some of the adverse effects of atrial fibrillation are felt to be due to the irregularly irregular rhythm, and this regularized rhythm may in part contribute to the beneficial effects of this procedure.
b. Rhythm Control
            As discussed above, the primary reason to choose a rhythm control strategy is for symptom control. There are several options to help an atrial fibrillation patient both achieve and maintain normal sinus rhythm (ie, rhythm control).
i. Cardioversion
Cardioversion simply refers to the process of converting from an abnormal rhythm to a normal rhythm. For atrial fibrillation, this can be achieved in 3 ways: 1) It can be spontaneous (as in the case of individuals with paroxysmal atrial fibrillation), 2) It can be done electrically, and 3) It can be done with a medicine.
            Electrical cardioversion is extremely effective in the short term and can restore sinus rhythm in the majority of patients.1 The problem is that atrial fibrillation often recurs and so maintaining sinus rhythm becomes an issue in and of itself. Electrical cardioversion is performed under deep sedation, with the patient made sleepy enough via medicines administered via an IV line so as not to feel the electric shock that is administered. Although the shock can be delivered using hand-held paddles, more commonly it is now done with 2 adhesive pads (typically one on the chest and one on the back). A shock of DC energy, typically somewhere between 100 and 360 joules is delivered. The procedure is generally very safe and well tolerated. Rarely, skin burns at the site of the adhesive pads or paddles can occur. Even more rarely, dangerous rhythms resulting from the electric shock can occur, but fortunately a second administration of electricity and/or medicines through the IV line are typically sufficient to take care of such an event. The procedure is generally done on an outpatient basis and the patient is usually sent home a few hours after arrival.
            Cardioversion with medicines is generally not as successful as using DC energy. However, the advantage is that the process is painless and therefore does not require any sedation. One advantageous application is called the “pill in the pocket” approach. In certain patients with very symptomatic and infrequent episodes of paroxysmal atrial fibrillation, an antiarrhythmic medicine that has the capability to pharmacologically cardiovert can be taken at home on an as needed basis. Although these agents are not 100% effective, those proportion of events that can be averted can often prevent emergency room visits and potentially, admissions to the hospital.12
Stroke prevention during and after cardioversion 
            One very important aspect of cardioversion is stroke prevention. Available evidence suggests that the time before and after a cardioversion is performed (regardless of the mode, whether spontaneous, electric, or pharmacologic) is a particularly high risk period for stroke.1 Strokes in atrial fibrillation are thought to occur due to the formation of a thrombus (or blood clot as above in Sections II and V) primarily in a part of the left atrium called the left atrial appendage. One concern is that, with the transition from atrial fibrillation to normal sinus rhythm, a preexisting clot can become dislodged. Just as important is the concern that a thrombus might form shortly after the cardioversion. Normally, the left atrial appendage contracts with great vigor, keeping blood moving and preventing blood from clotting inside of itself. With atrial fibrillation, the contracting function of the left atrial appendage can decrease. Of particular interest to anyone undergoing cardioversion, the left atrial appendage function can actually decrease further upon the transition from atrial fibrillation to normal sinus rhythm. The primary determinant of the severity and duration of what is called left atrial appendage “stunning” appears to rely on the duration of atrial fibrillation prior to the cardioversion. In fact, the generally accepted time frame after which this is felt to become a real concern is 48 hours; although there is no rigorous evidence to support this, cardioversion within 48 hours of the onset of atrial fibrillation is not generally thought to be associated with a substantial reduction in left atrial appendage function (and therefore not thought to be associated with a substantial risk for stroke).
            However, many scheduled cardioversions occur well after 48 hours of atrial fibrillation, and therefore 1 of 2 strategies can be employed to reduce the risk for stroke. First, the patient with atrial fibrillation can be treated with warfarin for 3 weeks prior to the cardioversion.1 Importantly, simply taking warfarin for this amount of time is insufficient. The warfarin must be maintained at a consistently therapeutic level, as determined by the international normalized ratio (INR) blood test during that period of time. Alternatively, a special ultrasound of the heart or echocardiogram can be performed to visualize the left atrial appendage in order to confirm the absence of a thrombus. The majority of echocardiograms are performed by placing an ultrasound transducer on the chest wall (called a transthoracic echocardiogram). However, the left atrial appendage is in the back of the heart, far away from the chest wall. It just so happens that the tube we use to swallow (the esophagus), connecting our mouth to our stomach, runs directly behind the left atrial appendage. Therefore, in order to visualize the left atrial appendage well, a transesophageal echocardiogram is performed, during which the patient is sedated and an ultrasound probe is advanced into the esophagus. If no left atrial appendage thrombus is observed, the cardioversion can be performed shortly, or immediately thereafter .1, 13 Regardless of whether 3 weeks of therapeutic warfarin or a transesophageal echodcardiogram is employed prior the procedure, all of these patients require at least 4 weeks of therapeutic anticoagulation to prevent stroke due to left atrial appendage stunning. Although warfarin is the ideal agent for this purpose, it often takes several days to reach a therapeutic level and therefore other agents (such as intravenous unfractionated heparin or subcutaneous low molecular weight heparin) are sometimes used to “bridge” the warfarin therapy.
ii. Antiarrhythmic medicines for the maintenance of normal sinus rhythm 
            Several antiarrhythmic agents can be used to help a person maintain sinus rhythm.1 Typically, these will be used in patients with paroxysmal atrial fibrillation or after cardioversion in patients with persistent atrial fibrillation. As noted above (Section VI.1.), the primary reason to use these agents are to treat symptoms only as there is no data to support their use for any long term benefit. These agents have traditionally been medicines that block certain ion channels responsible for different aspects of the heart’s normal conduction. Many are not well tolerated and only those mentioned in current guidelines will be briefly touched on here.1 In large part, by blocking certain ion channels, these antiarrhythmic drugs can also be proarrhythmic (in other words, they can actually increase the propensity to certain arrhythmias). As currently available agents are not specific to atrial tissue, the primary concern in using many of these agents involves the potential for dangerous (and even life threatening) proarrhythmia in the ventricles. Therefore, the choice of antiarrhythmic agents relies largely on the risk of proarrhythmia of the individual being treated, which itself depends on the presence and/ or nature of any underlying heart disease.
            In those with no underlying heart disease (as above, Section III., so-called lone atrial fibrillation patients), the risk of proarrhythmia is felt to be very low. First line agents for this group would include flecainide, propafenone, and sotalol. Although these agents are not the most potent, they can be very effective for many patients and are generally well tolerated. If these agents fail (or have adverse effects), either dofetilide or amiodarone can be considered. Dofetilide tends to be quite effective and is generally well tolerated; however, it can induce a dangerous ventricular rhythm called torsades de pointes, and patients (even those with no underlying heart disease) need to be rigorously monitored in the hospital for at least 3 days before it can be safely prescribed. Amiodarone is probably the most effective agent available for the maintenance of normal sinus rhythm, but it has multiple potential toxicities, including adverse effects involving the thyroid, lung, liver, skin, eyes, and nervous system, that classically occur when the medicine is used over a long period of time. Therefore, in general, the use of amiodarone should be avoided in young, otherwise healthy people unless absolutely needed.
            In those with coronary artery disease (or blocked coronary arteries) or any history of a myocardial infarction (or a heart attack), flecainide and propafenone are felt to be dangerous and should therefore be avoided.14 In those with heart failure due to weakened ventricles, only dofetilide and amiodarone are felt to be safe options. Finally, in those with very thick ventricles, amiodarone is the only agent that can be used safely.           
iii. Ablation of atrial fibrillation            
Recently, in an attempt to reproduce the great success of catheter ablation for the treatment of various arrhythmias (such as for AV nodal ablation described above in Section VI.1.a.ii), an interest in curing atrial fibrillation by catheter ablation is growing.15 This procedure involves placing multiple electrode catheters in the heart, and the focus of ablation is typically on the left atrium. Because the left atrium is not immediately accessible via the veins that drain into the right atrium, a transseptal approach is required:  approximately 20-30% of all adults will have a small hole in the septum between the right and left atrium that can be used, and, in the remainder, a needle is used to puncture the septum (in the area of a very thin membrane that is the remnant of the hole present during fetal development). A long sheath is then advanced over the long needle, the long needle is removed, and the ablation catheter is then maneuvered into the left atrium. Because the blood in the left atrium is contiguous with blood flow to the brain and other vital organs, heparin (an intravenous anticoagulant) is always infused immediately after left atrial access is obtained in order to avoid any potentially dangerous blood clot formation on the catheter (which, if dislodged, could result in a stroke). Although the exact technique by which curative left atrial ablation is performed varies somewhat from institution to institution, the general goal is to burn atrial tissue just outside the opening to the veins that empty blood into the left atrium from the lungs (the pulmonary veins). The general goal is to electrically isolate these pulmonary veins from the rest of the left atrium (Figure 6), and the proposed mechanisms behind the effectiveness of this strategy involve several concepts: first, there is some evidence that automatic focal electrical depolarizations arising from the pulmonary veins may be responsible for initiating atrial fibrillation in some people; by electrically isolating the veins, these impulses can not be propagated to the atrium and therefore theoretically not initiate atrial fibrillation. In addition, some believe that the series of burns delivered sufficiently disrupts the atrial substrate so as to disallow any of the reentrant wavelets from forming or sustaining (see Section II.). The vagal input to the atrium arises from several focal collections of nerves called ganglionic plexi, which happen to sit very near the openings of the pulmonary veins. There is some evidence that the effectiveness of the left atrial ablations depends largely on the ablation of these ganglia (something that may be accomplished while burning around each of the veins). Multiple explanations and techniques exist, but the procedure appears to be effective in approximately 60-80% of cases. Unfortunately, the risks of the procedure, although low (typically 4% or less), can be severe, including stroke, perforation of the heart, and very rarely, even death. Therefore, appropriate patient selection for this procedure is critical: young, otherwise healthy candidates with paroxysmal atrial fibrillation will likely benefit the most from this procedure with the lowest risk. The procedure continues to evolve, with techniques and equipment under constant development to improve the success and minimize the risk of the procedure, potentially allowing for the safe application of the procedure to a much broader patient population. For now, given that it is an invasive procedure which carries some risk, the general recommendation is that it be reserved for patients that are symptomatic despite medical therapy and have failed at least one antiarrhythmic agent (as described in Section VI.1.b.ii.).1, 15 In some circumstances, if a patient is very motivated to have the procedure and is felt to be a good candidate by his/her treating physicians (ie, felt to have a chance of high success and low risk), the procedure can be considered as first line therapy.

Figure 6. These panels demonstrate an example of one atrial fibrillation ablation technique. On the right panel, 2 catheters can be seen arising from a large vein (the inferior vena cava) below the heart, into and through the right atrium, across the interatrial septum (see text for method of crossing the septum) and into the left atrium. Of note, both catheters were introduced into the body via a vein in the leg. On the left panel, the ends of the catheters placed just outside the entrance of the left upper pulmonary vein are illustrated: the heated straight catheter delivers radiofrequency energy and thereby burns/ ablates atrial tissue; the circular catheter aids in recording electrical signals and can be helpful in determining when a structure (such as a pulmonary vein) has been “electrically isolated.” This figure was obtained with permission from Mr. David Criley at www.blaufuss.org.
           In general, until more data is available, the procedure should not be considered a substitution for warfarin therapy. As above (Section VI.1), some believe that with rigorous monitoring of the heart rhythm with wearable monitors for a month or more at a time, the absence of atrial fibrillation may provide enough evidence to safely discontinue warfarin therapy. Importantly, as below, the best candidates for the procedure (younger patients with lone atrial fibrillation), do not typically require warfarin therapy anyway. 

 2. Stroke Prevention

        The most effective agent to prevent stroke in atrial fibrillation patients is warfarin.1, 3 This medicine blocks the action of vitamin K, rendering the blood less prone to clot. Not surprisingly then, the primary adverse effect of warfarin is bleeding. In fact, dosing of warfarin involves a nearly constant effort to reach the balance between optimal stroke protection and minimal bleeding risk. Unfortunately, the potency of warfarin can be very unpredictable, from person to person and over time in the same person. Part of this inter-individual variability probably has something to do with different genes that can determine the way that warfarin is metabolized. Certainly, the potency of warfarin is determined in large part by diet: foods such as leafy green vegetables that contain more vitamin K will inhibit more of warfarin’s actions. Because of its unpredictable nature, the effect of warfarin must be measured regularly and the dose adjusted accordingly. A simple and readily available blood test measures the international normalized ratio (or INR), a measure of the blood’s clotting ability. A “normal” INR, for instance in someone not on warfarin, is approximately 1. A variety of large studies have shown that the optimal effectiveness of warfarin and the minimal bleeding risk in the setting of atrial fibrillation occur when the INR is between 2 and 3. Therefore, individuals on warfarin therapy must have regular blood tests, with a healthcare provider carefully reviewing the results and adjusting the dose of the medicine. Typically, tests and adjustments are more frequent during the initiation of therapy, and the goal is to require only monthly tests if a stable dose (with a stable INR between 2 and 3) can be achieved. Despite this, it is well recognized that even some of the most compliant and motivated patients will sometimes have INRs that are too low or too high. It is also important for patients on warfarin to avoid foods high in vitamin K and, perhaps more importantly, to maintain a consistent diet in vitamin K containing foods. Finally, warfarin patients must also be aware of interactions with other drugs (which are, unfortunately, quite common).  Healthcare providers that prescribe and adjust warfarin can often be very helpful in counseling patients regarding these important interactions.
            The only other alternative to warfarin that has thus far proven safe and effective in preventing strokes in the setting of atrial fibrillation is a daily aspirin. Although bleeding risk with aspirin is in general less than that of warfarin, bleeding does remain a concern (particularly gastrointestinal bleeding). It is also clear that aspirin is not as effective as warfarin in preventing stroke. Determining who should receive what anti-stroke treatment depends on a physician’s assessment of the person’s stroke risk.
            Because warfarin therapy involves some risk, the appropriate candidates for warfarin therapy are those in whom the benefit exceeds the risk. For example, in young lone atrial fibrillation patients (specifically, those younger than 60 with no heart or lung disease including hypertension), the risk of stroke is very low- warfarin is actually contraindicated in these patients because the risk of the therapy outweighs the very small benefit it might provide. Determining the additional risk factors for stroke in atrial fibrillation patients is critical in determining whether or not warfarin therapy is appropriate: increasing age (≥ 75 years old is a cut-off that has been used), hypertension (even if effectively treated), weakened ventricles (or heart failure), diabetes, and a history of stroke or transient ischemic attack (TIA)are the most important risk factors. According to current guidelines, if a patient has 2 or more of these risk factors, they should be on warfarin.116 There are other risk factors, including female gender and underlying coronary artery disease, that may confer some extra risk and may aid in the decision making. If a person only has one of these risk factors, either warfarin or aspirin therapy is felt to be acceptable and should be determined by the physician’s assessment of bleeding and stroke risks. Of note, a previous history of stroke by itself is felt to be a particularly important risk factor.

Conclusions

            Atrial fibrillation is an extremely common heart rhythm disorder with multiple potential adverse consequences. While the diagnosis is straightforward, the management can be quite complex and requires a comprehensive evaluation of a particular patient’s risks.  Excellent reference articles include the following: a comprehensive review contained in the joint American College of Cardiology/ American Heart Association/ European Society of Cardiology guidelines.1 The full text is available free at: http://circ.ahajournals.org/cgi/content/full/114/7/e257
                An outstanding review regarding curative atrial fibrillation ablation is provided by the Heart Rhythm Society.15 The full text can be found on the webpage: http://www.hrsonline.org/Education/AFib360/Guidelines/index.cfm
            Finally, a clearly illustrated and animated program on electrocardiograms and supraventricular arrhythmias is available at http://www.blaufuss.org/ (note: the complimentary tutorial can be found at the bottom of this home page, with animation on normal conduction, atrial fibrillation, and ablation under “SVT tutorial.”).
To learn more about the treatment of arrhythmias at the University of California, San Francisco (UCSF), please visit www.ucsfhealth.org/arrhythmia

References

1.         Fuster V, Ryden LE, Cannom DS, Crijns HJ, Curtis AB, Ellenbogen KA, et al. ACC/AHA/ESC 2006 guidelines for the management of patients with atrial fibrillation--executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the European Society of Cardiology Committee for Practice Guidelines (Writing Committee to Revise the 2001 Guidelines for the Management of Patients With Atrial Fibrillation). J Am Coll Cardiol 2006;48(4):854-906.
2.         Naito M, David D, Michelson EL, Schaffenburg M, Dreifus LS. The hemodynamic consequences of cardiac arrhythmias: evaluation of the relative roles of abnormal atrioventricular sequencing, irregularity of ventricular rhythm and atrial fibrillation in a canine model. Am Heart J 1983;106(2):284-91.
3.         Singer DE, Albers GW, Dalen JE, Go AS, Halperin JL, Manning WJ. Antithrombotic therapy in atrial fibrillation: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004;126(3 Suppl):429S-456S.
4.         Chugh SS, Blackshear JL, Shen WK, Hammill SC, Gersh BJ. Epidemiology and natural history of atrial fibrillation: clinical implications. J Am Coll Cardiol 2001;37(2):371-8.
5.         Jahangir A, Lee V, Friedman PA, Trusty JM, Hodge DO, Kopecky SL, et al. Long-term progression and outcomes with aging in patients with lone atrial fibrillation: a 30-year follow-up study. Circulation 2007;115(24):3050-6.
6.         Van Gelder IC, Hagens VE, Bosker HA, Kingma JH, Kamp O, Kingma T, et al. A comparison of rate control and rhythm control in patients with recurrent persistent atrial fibrillation. N Engl J Med 2002;347(23):1834-40.
7.         Wyse DG, Waldo AL, DiMarco JP, Domanski MJ, Rosenberg Y, Schron EB, et al. A comparison of rate control and rhythm control in patients with atrial fibrillation. N Engl J Med 2002;347(23):1825-33.
8.         Corley SD, Epstein AE, DiMarco JP, Domanski MJ, Geller N, Greene HL, et al. Relationships between sinus rhythm, treatment, and survival in the Atrial Fibrillation Follow-Up Investigation of Rhythm Management (AFFIRM) Study. Circulation 2004;109(12):1509-13.
9.         Page RL, Wilkinson WE, Clair WK, McCarthy EA, Pritchett EL. Asymptomatic arrhythmias in patients with symptomatic paroxysmal atrial fibrillation and paroxysmal supraventricular tachycardia. Circulation 1994;89(1):224-7.
10.       Gregoratos G, Abrams J, Epstein AE, Freedman RA, Hayes DL, Hlatky MA, et al. ACC/AHA/NASPE 2002 guideline update for implantation of cardiac pacemakers and antiarrhythmia devices: summary article: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (ACC/AHA/NASPE Committee to Update the 1998 Pacemaker Guidelines). Circulation 2002;106(16):2145-61.
11.       Wood MA, Brown-Mahoney C, Kay GN, Ellenbogen KA. Clinical outcomes after ablation and pacing therapy for atrial fibrillation : a meta-analysis. Circulation 2000;101(10):1138-44.
12.       Alboni P, Botto GL, Baldi N, Luzi M, Russo V, Gianfranchi L, et al. Outpatient treatment of recent-onset atrial fibrillation with the "pill-in-the-pocket" approach. N Engl J Med 2004;351(23):2384-91.
13.       Klein AL, Grimm RA, Murray RD, Apperson-Hansen C, Asinger RW, Black IW, et al. Use of transesophageal echocardiography to guide cardioversion in patients with atrial fibrillation. N Engl J Med 2001;344(19):1411-20.
14.       Preliminary report: effect of encainide and flecainide on mortality in a randomized trial of arrhythmia suppression after myocardial infarction. The Cardiac Arrhythmia Suppression Trial (CAST) Investigators. N Engl J Med 1989;321(6):406-12.
15.       Calkins H, Brugada J, Packer DL, Cappato R, Chen SA, Crijns HJ, et al. HRS/EHRA/ECAS expert Consensus Statement on catheter and surgical ablation of atrial fibrillation: recommendations for personnel, policy, procedures and follow-up. A report of the Heart Rhythm Society (HRS) Task Force on catheter and surgical ablation of atrial fibrillation. Heart Rhythm 2007;4(6):816-61.
16.       Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW, Radford MJ. Validation of clinical classification schemes for predicting stroke: results from the National Registry of Atrial Fibrillation. JAMA 2001;285(22):2864-70.