Atrial fibrillation following cardiac surgery is very common, with most series reporting an incidence of between 30% and 60%.^1,² The peak incidence is on the second postoperative day. Postoperative atrial fibrillation is not benign; it prolongs ICU and hospital stay, increases costs, and is associated with an increased incidence of stroke.²

The arrhythmia is characterized by the abrupt onset of an irregularly irregular narrow-complex tachycardia (Fig. 21-3), often with varying R wave amplitude. Heart rate varies widely, depending on adrenergic tone, but is typically between 120 and 160 beats per minute. In cardiac surgery patients, the onset of atrial fibrillation may cause marked hypotension.

Figure 21.3 Rhythm strip showing atrial fibrillation. The QRS rate is irregularly irregular, with a rate of approximately 110 beats per minute. Variable atrial activity is present but no clear P waves are seen. (Standard sweep speed is 25 mμ/sec.)

Pathophysiology

Atrial fibrillation was regarded for a long time as a classic example of random reentry due to multiple wandering wavelets.¹ However, it is now recognized that in many patients the atria are only passively involved and that the initiating trigger is rapid focal discharges from cells located in the muscle sleeves of the pulmonary veins.³

Sustained atrial fibrillation leads to mechanical and electrical remodeling of the atria, with loss of contractility (atrial stunning) and shortening of refractory periods. Atrial stunning persists for several weeks after reestablishment of sinus rhythm, increasing the risk for atrial thrombus formation, especially in the left atrial appendage. The shortening of atrial refractory periods results in reinforcement of the arrhythmia the longer it remains untreated.⁴ Furthermore, patients are very vulnerable to reinduction of atrial fibrillation immediately after reversion to sinus rhythm. For these reasons, after successful cardioversion of atrial fibrillation of more than 48 hours’ duration, anticoagulation and antiarrhythmic therapy should continue for at least 30 days.⁵

Risk Factors

The principal risk factor for atrial fibrillation after cardiac surgery is increased age.² Previous atrial fibrillation, mitral and aortic surgery, atrial enlargement, and withdrawal of β blockers are also important risk factors. Other factors associated with the development of atrial fibrillation include the requirement for inotropes or vasopressors, hypomagnesemia, prolonged cardiopulmonary bypass, and impaired ventricular function. Acute atrial stretch due to impaired ventricular function or tamponade may precipitate atrial fibrillation. Thus, atrial fibrillation may be both a cause and a consequence of hemodynamic instability.

Prophylaxis

β Blockers (metoprolol, carvedilol, and sotalol) and amiodarone reduce the incidence of atrial fibrillation by as much as 50%.⁶^–⁹ The combination of amiodarone and a β blocker appears to be more effective than either agent alone.¹⁰ Other drugs such as magnesium and diltiazem may also be useful for prophylaxis but data are inconclusive.⁹ Standard (right atrial) pacing may have a small protective effect against atrial fibrillation, but the benefit is greater if simultaneous left and right atrial pacing is used.^11,¹² Digoxin is not indicated for prophylaxis of postoperative atrial fibrillation.^7,⁹

Prophylaxis against atrial fibrillation should include avoiding β-blocker withdrawal in patients chronically treated with them. For patients at increased risk for atrial fibrillation, routine prophylaxis with a β blocker or amiodarone is appropriate.⁹ Prophylactic biatrial pacing should also be considered in high-risk patients.¹²

The preoperative administration of 600 mg per day of amiodarone for 1 week prior to surgery has been shown to be effective ⁸ but may not be feasible. A practical alternative is intraoperative dosing of 5 mg/kg intravenously followed by oral treatment of 400 mg three times daily for 2 days, 200 mg twice daily for 2 days, and then 200 mg daily until hospital discharge. Amiodarone should be withheld in the presence of sick sinus syndrome, bradycardia (heart rate <50 beats per minute), or marked first-degree AV block (PR interval >0.30 seconds). In patients with paroxysmal atrial fibrillation who are chronically treated with amiodarone, low-dose oral sotalol (40 to 80 mg twice daily) may be commenced postoperatively, with careful monitoring for excessive bradycardia or QT prolongation. When atrial fibrillation is associated with planned mitral valve surgery, a surgical maze procedure should be considered (see Chapter 10).

Treatment

Atrial fibrillation that is associated with hemodynamic compromise should be treated aggressively. The success rate for primary cardioversion is low—less than 20% in one study.¹³ The success rate of electrical cardioversion may be increased by the use of biphasic energy (with an energy level of 200 joules) and the prior correction of hypokalemia and hypomagnesemia. If atrial pressures are elevated, an echocardiogram should be considered to rule out ventricular dysfunction and pericardial tamponade. If cardioversion is unsuccessful, it may be repeated following intravenous amiodarone (a bolus dose of 5 mg/kg over 30 minutes or an infusion of 2 mg/min for 4 hours). Amiodarone provides acute control of heart rate through its β-blocker actions and may itself result in pharmacologic cardioversion. Following successful cardioversion, an infusion of amiodarone (1 mg/min) should be administered until the patient is able to take oral amiodarone or a β blocker. Atrial pacing (AAI mode at 80 to 90 beats per minute for 24 to 48 hours) should be instituted to minimize the chance of recurrence. Intravenous sotalol or ibutilide may be used as an alternative in patients with good left ventricular function.¹⁴ If, despite these measures, the patient remains in rapid atrial fibrillation, pharmacologic control of heart rate may be achieved by an intravenous β blocker or diltiazem.¹⁵ In patients with unstable hemodynamics, intravenous digoxin may be tried, but it has limited efficacy in the setting of heightened adrenergic tone.¹⁵

In extubated patients who are not hemodynamically compromised it is reasonable to attempt pharmacologic cardioversion initially with intravenous amiodarone or a β blocker. If atrial fibrillation persists at 24 hours, the patient should be electively cardioverted (200 joules biphasic energy) followed by atrial pacing. If cardioversion is unsuccessful or atrial fibrillation is recurrent, pharmacologic rate control may be achieved with digoxin in combination with either a β blocker or diltiazem.

Pharmacologic treatment for atrial fibrillation should be continued for 4 to 6 weeks following surgery, even if the patient reverts to sinus rhythm.¹⁵ If atrial fibrillation persists at 6 weeks following surgery, a choice should be made between repeat cardioversion and permanent rate control and anticoagulation. The latter strategy is favored by the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) study, which did not show a better outcome for older patients with attempted maintenance of sinus rhythm by antiarrhythmic drugs.¹⁶

Cardioversion Guided by Transesophageal Echocardiography

Cardioversion should not be performed without the guidance of transesophageal echocardiography in patients who have experienced atrial fibrillation for more than 48 hours and are not undergoing therapeutic anticoagulation. This is so because of the risk for systemic embolization of intracardiac thrombi.¹⁷ If left atrial thrombus is present on transesophageal echocardiography (TEE) examination, the patient should be given anticoagulants with heparin and warfarin, and cardioversion should be performed 4 weeks later if repeat (TEE) shows resolution of the thrombus. If thrombus is not present, cardioversion may proceed immediately, after which the patient should undergo anticoagulation with warfarin for 4 to 6 weeks (see subsequent material).

Anticoagulation for Atrial Fibrillation

If there are no contraindications, patients with sustained or paroxysmal atrial fibrillation for more than 24 to 48 hours after cardiac surgery should be given an anticoagulant with warfarin, aiming for an International Normalised Ratio (INR, see Chapter 30) between 2 and 3. Warfarin should be continued for 4 to 6 weeks following surgery, at which time, if the patient is in sinus rhythm, it may be stopped.⁵ However, if there are factors predictive of recurrence, such as a history of paroxysmal atrial fibrillation, a left atrial diameter greater than 5 cm, or mitral valve surgery, warfarin should be continued.

Patients at increased risk for thromboembolic complications, notably those with documented atrial thrombus, a history of systemic embolization, or a prosthetic mitral or aortic valve, should also be given anticoagulants with unfractionated heparin, which should be continued until the patient has a therapeutic INR.⁵ Compared with warfarin, aspirin is less effective in reducing stroke frequency in patients with atrial fibrillation,^18,¹⁹ and it is therefore not usually an acceptable alternative.

Nonpharmacologic Treatment of Atrial Fibrillation

A number of catheter-based techniques involving isolation of the triggering foci around the pulmonary veins and ablation of fractionated atrial potentials have been developed for the treatment of chronic atrial fibrillation.^20,²¹ In older patients, radiofrequency ablation of the bundle of His with permanent pacemaker implantation provides good symptom relief, although the need for anticoagulation remains. These techniques have a minimal role in the management of postoperative atrial fibrillation.

Atrial Flutter

Risk factors for atrial flutter are similar to those of atrial fibrillation, although atrial flutter is far less common than atrial fibrillation.

Atrial flutter is usually caused by a reentry circuit around the tricuspid valve with an area of slow conduction near the orifice of the coronary sinus, which results in abnormal flutter waves on the ECG that have a characteristic sawtooth pattern at a rate of between 250 and 330 per minute. Because of the long refractory period of the AV node, usually only a proportion of the flutter waves are conducted to the ventricles; typically in a ratio of 2:1 (Fig. 21-4), resulting in a heart rate of about 150. Higher ratios of AV block (e.g., 3:1 or 4:1) resulting in slower heart rates are also seen (Fig. 21-5). Postoperative atrial flutter typically has 2:1 conduction. If antiarrhythmic drugs have been given as part of routine prophylaxis, the ventricular rate with 2:1 block may be less (125 to 150 beats per minute). Atrial flutter with 2:1 AV block can be difficult to distinguish from other, regular narrow-complex tachycardias because one of the flutter waves may be concealed within the QRS complex; the rhythm may be mistaken for sinus tachycardia with first-degree AV block. Negative flutter waves on the inferior ECG leads and a fixed heart rate of 140 to 150 per minute usually confirm the diagnosis, but if doubt persists, an atrial ECG should be performed if possible (Fig. 21-6). Alternatively, adenosine (6 or 12 mg) as a rapid intravenous bolus will cause temporary AV block, exposing the characteristic flutter waves.

Figure 21.4 Rhythm strip showing atrial flutter with predominantly 2:1 conduction. After the first complete QRS complex, a 1-second period of complete AV block is seen, revealing the characteristic sawtooth pattern of the atrial flutter. After that, a period of 2:1 conduction is illustrated; arrows mark the approximate positions of the flutter waves.

Figure 21.5 Rhythm strip showing atrial flutter with 4:1 conduction. Three clear flutter waves are seen between each QRS complex; the fourth is partially concealed within the QRS complex.

Figure 21.6 Atrial ECG showing atrial flutter with 2:1 conduction. The sweep speed is set at 50 mm per second. Lead I (bipolar atrial lead) captures only the signal of the flutter wave. Leads II and III capture the signal of the QRS (the larger amplitude deflection) in addition to the signal of the flutter waves (arrows in lead III).

Treatment

If atrial pacing wires are in place, overdrive pacing (see subsequent material) is the treatment of choice for acute atrial flutter. If overdrive pacing is unsuccessful, electrical cardioversion with 200 joules biphasic energy is usually effective. Hypokalemia and hypomagnesemia should be corrected prior to cardioversion. Occasionally, atrial flutter is resistant to overdrive pacing and electrical cardioversion. In this situation, intravenous amiodarone followed by a further attempt at electrical cardioversion is indicated. Attempts to control heart rate in patients with atrial flutter are often ineffective because the usual 2:1 AV conduction block is difficult to increase.

The risks for thromboembolism with atrial flutter are no different from those with atrial fibrillation, and the same indications for anticoagulation and cardioversion guided by TEE apply. Typical atrial flutter can be permanently prevented at the time of cardiac surgery by creating a surgical or radiofrequency scar line between the tricuspid annulus and the orifice of the inferior vena cava, extending from endocardium to epicardium (see Fig. 10-8).²²

Technique of Overdrive Pacing for Atrial Flutter

Within a reentry circuit, at any point in time a portion of the circuit is not being activated. This is the distance between the advancing head of the circulating wave front and the tail of depolarized cells; it is called the excitable gap. Pacing impulses can invade the circuit via this gap and, by colliding with the arrhythmia wave front, eliminate the arrhythmia (Fig. 21-7). This technique is known as overdrive pacing and is useful for certain reentry arrhythmias, including ventricular tachycardia and AV nodal reentry tachycardia, but it is particularly useful for atrial flutter.

Figure 21.7 Overdrive pacing. A, A reentrant circuit. The wave front advancing around the circuit leaves in its wake cells that are initially absolutely, and later relatively, refractory to depolarization. The remaining cells have fully repolarized and are capable of depolarization again. These cells constitute the “excitable gap” in the reentrant circuit. B, A high-frequency pacing source has been introduced; it promotes a second wave front that collides with the reentrant wave front and obliterates it at point X (marked). The wave front generated by the pacing source does not progress down the other limb of the circuit because the cells within it are in a refractory state (the “relative refractory tail”).

When two atrial epicardial wires are in situ, each wire should be tested to confirm that it is recording only an atrial ECG and to measure the atrial rate. The pacing pulse width should be increased to 2 ms and pacing begun at 20 mA at 100 beats per minute to confirm the absence of ventricular capture. The pacing rate should be increased to 20 beats faster than the intrinsic atrial rate (typical atrial rates are 250 to 280 beats per minute but may be as high as 330 beats per minute), and the ECG should be observed to confirm atrial capture. After 30 seconds, the pacing rate should be increased by a further 20 beats per minute. Atrial capture is confirmed by: (1) an increase in heart rate as the pacing rate is increased; (2) a subsequent abrupt fall in heart rate as the AV conduction ratio increases (from 2:1 to 3:1 or 4:1); and (3) a constant relationship between the pacing spikes and the flutter waves. Pacing is abruptly stopped after 1 to 2 minutes of atrial capture, which typically results in the establishment of sinus rhythm. If sinus rhythm is not present, the process should be repeated after reversal of atrial lead polarity. If it is still unsuccessful, deliberate induction of atrial fibrillation should be attempted by burst pacing at rates of 600 per minute for 30 seconds or until atrial fibrillation ensues. Pacing-induced atrial fibrillation is typically unstable and frequently reverts spontaneously to sinus rhythm, though reorganization to atrial flutter is possible.

Accelerated Junctional Rhythm

Accelerated junctional rhythm (nodal tachycardia) is a regular narrow-complex tachycardia in which no P wave can be seen preceding the QRS complex. The tachycardia typically develops gradually (warm up), slowly increasing up to a heart rate of 110 to 150 beats per minute. Retrograde-conducted P waves that follow the QRS are visible on an atrial ECG and may be seen on a standard rhythm strip (Fig. 21-8). Depending on the ventriculoatrial interval, regular cannon A waves may be visible on the central venous pressure (CVP) trace (see Chapter 8). Occasionally, the atrial ECG demonstrates AV dissociation, in which case irregular cannon waves are seen instead.

Figure 21.8 Rhythm strip showing accelerated junctional rhythm. The heart rate is approximately 95 beats per minute; the retrograde-conducted negative P waves (arrows) are seen just after the QRS complex within the inverted T wave.

The mechanism of accelerated junctional rhythm is either enhanced or abnormal automaticity, and pacing or cardioversion cannot terminate it. Accelerated junctional rhythm is virtually always a transient phenomenon in the setting of heightened adrenergic tone. If treatment is necessary, an intravenous β blocker is appropriate.

A malignant variation of accelerated junctional rhythm is junctional ectopic tachycardia (JET). JET occurs almost exclusively in neonates or young children undergoing major congenital cardiac surgery. JET is associated with marked tachycardia (>180 beats/min), important hemodynamic compromise, and increased mortality rates. The treatment is sedation, paralysis, cooling, and intravenous amiodarone.

Paroxysmal Supraventricular Tachycardia

Paroxysmal supraventricular tachycardia is uncommon as a de novo arrhythmia in the ICU setting; most instances occur in patients with previously recognized episodes. The two most common mechanisms are AV node reentry and reentry due to an accessory pathway. Less common are ectopic atrial tachycardias.

AV Node Reentry Tachycardia

Reentry within the AV node (“supraventricular tachycardia”) occurs as a consequence of abnormal nodal architecture, in which there are two routes across the AV node from the atria to the His bundle, a fast and a slow pathway.²³ A reentry circuit is created by antegrade block across the fast pathway, activation across the slow pathway that continues on to the ventricles, and retrograde activation back to the atria via the fast pathway.

It presents with the abrupt onset of a regular narrow-complex tachycardia with a typical rate of 150 to 180 beats per minute (Fig. 21-9). On the atrial ECG, retrograde-conducted P waves may be seen within the QRS and, depending on the ventriculoatrial interval, regular cannon A waves may be seen on the CVP trace.

Figure 21.9 Rhythm strip showing AV nodal reentry tachycardia. The heart rate is approximately 175 beats per minute, and no atrial activation is visible.

AV node reentry tachycardia is usually a nuisance rhythm, and it can be terminated easily by overdrive pacing (if temporary wires are in situ). Alternatively, intravenous adenosine 6 mg, given centrally as a bolus injection and increased if ineffective to 12 mg, is highly successful in terminating this rhythm. For recurrent AV node reentry tachycardia, intravenous amiodarone is useful. Oral sotalol 80 mg twice daily is a good choice for ongoing prophylaxis. Most patients who require long-term medication are candidates for radiofrequency ablation. AV node reentry tachycardia may be mistaken for rapid accelerated junctional rhythm and may be indistinguishable from it on the atrial ECG. AV node reentry is suggested by abrupt onset, a typically faster heart rate, and the ability to terminate the rhythm with overdrive pacing.

Accessory Pathways

An accessory pathway is usually a strand of atrial myocardium joining the atrium to the ventricle at the level of the tricuspid or mitral valve annulus. The most common type is associated with Wolff-Parkinson-White syndrome.

If conduction via the accessory pathway is faster than through the AV node, early (pre-) excitation of the ventricles occurs, resulting in a short PR interval and a delta wave on the resting ECG (Fig. 21-10). When conduction from atria to ventricles across the accessory pathway does not occur (so-called concealed accessory pathway), the resting ECG does not demonstrate preexcitation.

Figure 21.10 Rhythm strip showing preexcitation on the resting ECG in a patient with Wolff-Parkinson-White syndrome. The P wave (P) is labeled; the PR interval is short at approximately 0.08 seconds (2 small squares); and a delta wave (D) is seen within the QRS complex, effectively broadening it.

Patients with accessory pathways are at risk for paroxysmal supraventricular tachycardias. The most common arrhythmia associated with the Wolff-Parkinson-White syndrome is orthodromic tachycardia, in which antegrade conduction occurs via the AV node and retrograde conduction back to the atria occurs across the accessory pathway. This results in a narrow-complex regular tachycardia similar to AV node reentry tachycardia—and it responds to similar treatment. Less commonly, antidromic tachycardia occurs; in this case antegrade conduction occurs across the accessory pathway and retrograde conduction progresses up the normal conducting system. This results in a broad-complex tachycardia that is indistinguishable from ventricular tachycardia.²⁴ This is best treated by electrical cardioversion. Atrial fibrillation may be conducted across the accessory pathway at very fast rates, resulting in a broad-complex tachycardia that resembles ventricular tachycardia but is usually irregular.

Digoxin, β blockers, and calcium antagonists can potentially augment antegrade conduction across the accessory pathway and are therefore contraindicated in patients with preexcitation on a resting ECG.

Ectopic Atrial Tachycardias

Atrial tachycardias are narrow-complex, regular tachycardias in which each QRS complex is preceded by an abnormal P wave. Of the atrial tachycardias, only multifocal atrial tachycardia is likely to be encountered in the ICU.

Multifocal atrial tachycardia is an irregular narrow-complex tachycardia in which three or more atrial morphologies occur. It is typically associated with critical illness, particularly illness associated with a respiratory cause. It is best managed by correction of the underlying critical illness. In addition, magnesium, amiodarone, and β blockers may be used. Cardioversion is ineffective, and digoxin may worsen the tachycardia.

Atrial Premature Beats

Atrial premature beats are ectopic beats that originate in the atria. They have a premature and abnormal-looking P wave followed by a normal QRS complex. Atrial premature beats are associated with an incomplete compensatory pause, meaning that the interval between the preceding and following sinus beats is less than two complete cycles. If an atrial premature beat follows every sinus beat, atrial bigeminy is said to exist. If an atrial premature beat occurs early, it may not be conducted because the AV node is still refractory, and it appears as an atrial pause. Atrial premature beats can initiate reentrant atrial tachycardias (usually atrial flutter or fibrillation) but do not require any specific antiarrhythmic therapy.

BROAD-COMPLEX TACHYCARDIAS

Broad-complex tachycardia may be defined as a heart rate greater than 100 beats per minute and a QRS duration of more than 0.12 seconds (3 small squares on the ECG). Broad-complex tachycardia may be due to ventricular tachycardia, supraventricular tachycardia with aberrant conduction (i.e., bundle branch block), or conduction across an accessory pathway (e.g., antidromic tachycardia).

Ventricular Tachycardia Versus Supraventricular Tachycardia

The great majority of broad-complex tachycardias are ventricular tachycardias. It is essential to start with this mindset and disregard the common tendency to assume that if the patient has a stable blood pressure, the tachycardia must be a supraventricular tachycardia with aberrant conduction. In the presence of heart disease, broad-complex tachycardia should always be managed as ventricular tachycardia.

If there is important hemodynamic compromise, emergency cardioversion is required (with general anesthesia if the patient is conscious). In all other circumstances, the first step is to record a 12-lead ECG, including a long lead II rhythm strip. This ECG may then be compared with previous ECGs to help differentiate between ventricular and supraventricular origin.

If a previous ECG demonstrates bundle branch block (see subsequent material) with the same morphology as the tachycardia, supraventricular tachycardia is highly likely—unfortunately, aberrant conduction is often present only during tachycardia, so the absence of a bundle branch pattern on a previous ECG does not rule out a supraventricular origin. If a previous ECG shows ventricular extrasystoles that have the same morphology (lead for lead) as the tachycardia, ventricular tachycardia is diagnosed (Fig. 21-11). If a delta wave is present on a previous ECG, and the initial QRS vectors during tachycardia are an exact match, supraventricular conduction over an accessory pathway is probable; irregular RR intervals strongly suggest atrial fibrillation. Additional supporting information that is suggestive of ventricular tachycardia includes:

• “Rabbit ear” morphology in V1 where left is greater than right (Fig 21-12)

• QS complex in V6 (see Fig. 21-12)

• Left axis deviation, “northwest” axis (180 to 270 degrees), or indeterminate axis (see Fig. 21-12)

• Ventricular concordance; i.e., where all complexes are either positive or negative across the ventricular leads

• QRS duration greater than 0.14 seconds (>3.5 small squares)

• Fusion beats; i.e., the presence of a QRS complex with hybrid morphology in between that of a normal complex and a tachycardia complex (Fig. 21-13)

• Capture beats; i.e., the presence of a normal QRS complex in between tachycardia complexes (see Fig. 21-13)

• Evidence of AV dissociation on the 12-lead or atrial ECG; i.e., P waves that are independent of the QRS complexes.

Figure 21.11 Ventricular tachycardia. A, Extrasystoles on the resting 12-lead ECG have the same morphology, lead for lead, as the broad-complex tachyarrhythmia seen in B, indicating that it is ventricular tachycardia.

Figure 21.12 Ventricular tachycardia. The following features can be seen: (1) QRS duration greater than 140 ms; (2) left > right “rabbit ear” morphology in lead V1 (arrows); QS in lead V6, northwest axis. These are diagnostic of ventricular tachycardia.

Figure 21.13 Ventricular tachycardia rhythm strip showing atrioventricular dissociation. The P waves are shown (arrows). A fusion beat (F) and a capture beat (C) are seen; they indicate ventricular tachycardia (see text for details).

Ventricular Tachycardia

Ventricular tachycardia may be monomorphic (see Figs. 21-11 and 21-13), in which all the QRS complexes are of similar amplitude, or polymorphic, in which the QRS complexes are of variable amplitude (Fig. 21-14).

Figure 21.14 Rhythm strip showing polymorphic ventricular tachycardia with typical long-short initiation (see text for details). S, sinus beat; VPB, ventricular premature beat.

Monomorphic Ventricular Tachycardia

Monomorphic ventricular tachycardia most often has a right bundle branch block (RBBB) morphology (see subsequent material under Bundle Branch Block and Fascicular Block), reflecting left ventricular origin. When a left bundle branch block (LBBB) morphology is present, septal or right ventricular involvement is likely and indicates the need for subsequent further investigation (see subsequent material under Follow-up Management). Both types of ventricular tachycardia occur in critically unwell patients with impaired cardiac function, but monomorphic ventricular tachycardia is not usually a manifestation of ischemia; it is typically associated with scarring resulting from a previous myocardial infarction.

Monomorphic ventricular tachycardia is a more organized rhythm than the polymorphic form, and patients may maintain a reasonable hemodynamic state. In the absence of hypotension, monomorphic ventricular tachycardia can be treated with intravenous sotalol (1 mg/kg to a maximum of 100 mg) or amiodarone (5 mg/kg). Intravenous sotalol is effective but it may cause marked hypotension. Lidocaine is usually safe but is only weakly effective. Slow ventricular tachycardia with maintained blood pressure can also be terminated by a 6- to 12-beat burst of ventricular pacing at a rate 10 to 40 beats faster than the intrinsic rhythm. Pacing-induced acceleration or ventricular fibrillation may occur, so a defibrillator should always be immediately available. In patients with ventricular tachycardia who are hypotensive or have impaired consciousness, immediate electrical cardioversion is indicated.

Polymorphic Ventricular Tachycardia

Polymorphic ventricular tachycardia is a rapid ventricular tachycardia with a characteristic variable morphology and sinusoidal variation in amplitude. Polymorphic ventricular tachycardia with an alternating QRS morphology is often associated with prolongation of the QT interval during sinus rhythm, in which case it is known as torsades de pointes. Torsades de pointes is thought to be initiated by delayed after-depolarizations and maintained by reentry. On the ECG rhythm strip, a ventricular premature beat followed by a long pause followed by a conducted QRS complex and a ventricular premature beat, which initiates the arrhythmia, is commonly seen (the so-called long-short initiation; see Fig. 21-14). Torsades de pointes is commonly recurrent. The ECG appearances of polymorphic ventricular tachycardia may be difficult to distinguish from “coarse” ventricular fibrillation (Fig. 21-15). Factors that worsen QT prolongation increase the risk for developing torsades de pointes; namely, bradycardia and AV block, hypokalemia, and antiarrhythmic drugs with class III activity. Polymorphic ventricular tachycardia may also occur in the absence of QT prolongation, most commonly in the setting of acute myocardial ischemia or infarction.

Figure 21.15 Rhythm strip showing coarse ventricular fibrillation. No discrete QRS complexes are visible.

Polymorphic ventricular tachycardia deriving from any cause is typically poorly tolerated and may degenerate into ventricular fibrillation. Immediate electrical cardioversion is usually necessary. In addition to cardioversion, treatment of torsades de pointes includes intravenous magnesium sulfate (1 to 2 g slowly over 1 to 2 minutes),²⁵ pacing to a rate of at least 90 beats per minute, and correction of hypokalemia (aiming for a potassium concentration >4.5 mmol/l). Class III antiarrhythmic drugs such as sotalol and amiodarone are contraindicated, and agents such as lidocaine and β blockers are ineffective.

Polymorphic ventricular tachycardia that is associated with a normal resting QT interval may be managed in the same way as the monomorphic form. However, in practice the safest option is to administer magnesium and perform electrical cardioversion in all patients who present with recurrent polymorphic ventricular tachycardia, irrespective of the QT interval.

Recurrent Ventricular Tachycardia

Recurrent ventricular tachycardia is an emergency situation that is associated with high mortality rates. Left ventricular function is usually impaired, and the combined effects of tachycardia-related hypotension and temporary myocardial stunning resulting from repeated electrical cardioversion rapidly lead to irreversible hemodynamic decline. If possible, a 12-lead ECG of the tachycardia and of sinus rhythm should be obtained to confirm the diagnosis of ventricular tachycardia and to determine the QT interval.

Magnesium should be given empirically regardless of the QRS morphology because it is safe and sometimes highly effective. If torsades de pointes is excluded, amiodarone may be given as a bolus dose of 5 mg/kg followed by an infusion of 1 mg/min (see Table 3-6). The serum potassium should be increased to more than 5 mmol/l. If ventricular tachycardia persists, the addition of an intravenous β blocker may be highly effective. Occasionally, the addition of lidocaine to amiodarone may also be effective: the effect of two boluses of 75 mg 5 minutes apart, with the subsequent infusion of 2 mg/min can be tried; it can be discarded if not rapidly effective.

Recurrent ventricular tachycardia, particularly that which degenerates into ventricular fibrillation, is a common manifestation of acute ischemia or severely impaired ventricular function. Left ventricular function should be assessed by echocardiography. Hemodynamics should be supported by an intraaortic balloon pump and inotropic agents as required. If there is evidence of acute ischemia, emergency angiography followed by revascularization is indicated. Recurrent ventricular tachycardia after a revascularization procedure may indicate an acute coronary occlusion, and repeat revascularization should be considered. If the patient has a discrete anteroapical aneurysm, emergency aneurysmectomy and endocardial peel may be dramatically effective, but it is a very risky undertaking.

Measurement of the QT Interval

It is often hard to be certain where the T wave merges with the baseline. Consequently, measuring the QT interval is imprecise and is variable among different observers.²⁶ Lead II is a good choice for measurement. It may be helpful to draw a line along the downslope of the T wave to the point where it intersects the baseline taken as the end of the QT interval (see Fig. 8-4). The interval varies according to heart rate. As a consequence it is normalized as the QTc, most commonly by dividing the measured QT interval by the square root of the preceding RR interval. Conventional normal values are less than or equal to 0.44 seconds for males and less than or equal to 0.46 seconds for females. With torsades de pointes, the QTc is commonly more than 0.55 seconds. When the QT interval is prolonged, the following U wave is usually prominent too and often is fused with the T wave. If there is no clear separation, the combined QTU interval should be measured.

Follow-up Management of Ventricular Tachycardia

If ventricular tachycardia occurs solely in the first 24 hours after heart surgery in the context of myocardial dysfunction, there is usually no long-term predisposition to recurrence. For ventricular tachycardia occurring later than that, a strategy for future protection of the patient is required. This most often will be implantation of a cardioverter-defibrillator. In patients with ejection fractions greater than 40% and hemodynamically tolerated ventricular tachycardia, long-term sotalol or amiodarone is an acceptable alternative. In patients with LBBB morphology ventricular tachycardia, especially if left ventricular function appears relatively normal, imaging of the right ventricle to confirm or exclude arrhythmogenic right ventricular cardiomyopathy is indicated.²⁷

Nonsustained Ventricular Tachycardia

Nonsustained ventricular tachycardia is a common arrhythmia and although of prognostic importance when linked to heart disease, management is empiric rather than of proven advantage. As a rule, it reflects transient electrolyte disturbance, especially hypokalemia, in the immediate postoperative period. When nonsustained ventricular tachycardia persists after potassium repletion, patients should be examined for evidence of ventricular dysfunction, myocardial infarction, and acute ischemia. Persistent arrhythmias in patients with ejection fractions less than 40% and coronary artery disease are indications for an electrophysiologic study, looking for inducible sustained ventricular tachycardia.²⁸ Patients with inducible sustained ventricular tachycardia are candidates for implantable cardioverter-defibrillators.

Ventricular Fibrillation

Ventricular fibrillation is a macro reentrant arrhythmia in which there are no clearly formed QRS or T waves. The rhythm has a gradually undulating baseline and looks the same when viewed upside down. Initially, the appearances are similar to those of polymorphic ventricular tachycardia, with relatively coarse fibrillation waves (see Fig. 21-15), but over a period of minutes the amplitude of the waves becomes progressively smaller (fine ventricular fibrillation), until eventually asystole develops. Ventricular fibrillation is not associated with any cardiac output and thus defines a state of cardiac arrest. Treatment of ventricular fibrillation is outlined in Chapter 20. In any patient who develops recurrent ventricular fibrillation (or ventricular tachycardia that deteriorates into ventricular fibrillation), the presence of acute myocardial ischemia should be a considered. In this situation, emergency coronary revascularization may be a lifesaving treatment. If recurrent ventricular fibrillation is persistently initiated by the same premature ventricular contraction (constant morphology), an urgent electrophysiology consultation should be sought with a view to emergency ablation of the triggering focus.

Ventricular fibrillation is not uncommon after cardiac surgery. A single episode in the first 24 hours following surgery does not usually carry prognostic significance, particularly if it is not associated with myocardial ischemia. It is important to rule out an R-on-T phenomenon secondary to asynchronous ventricular pacing (see the subsequent material under Temporary Epicardial Pacing). Episodes of ventricular fibrillation after 24 to 48 hours require individual evaluation; in critically unwell patients (who subsequently recover), they may have no long-term significance. If episodes occur more than 48 hours after surgery, patients should be regarded as being at future risk for recurrence and should be evaluated for insertion of a cardioverter-defibrillator before discharge.

Ventricular Premature Beats

Ventricular premature beats are ectopic beats that originate in the ventricles. A typical ventricular premature beat occurs early—on or shortly after the T wave of the preceding sinus beat—and is associated with a wide QRS complex (Fig. 21-16). Ventricular premature beats are usually associated with a full compensatory pause, meaning that the interval between the preceding and following sinus beats is equal to two complete cycles (compare this with atrial premature beats, described earlier). If a ventricular premature beat follows every sinus beat, ventricular bigeminy is said to exist. When ventricular premature beats have different QRS morphologies they are assumed to arise from different foci and are commonly labeled multifocal (strictly speaking, they are multimorphic because different exits from a similar focus are also possible). When ventricular premature beats occur very early—with the R wave of the ectopic beat occurring on the T wave of the preceding beat—they can induce ventricular fibrillation. This is most likely to occur in the setting of myocardial infarction, hypokalemia, or QT prolongation. Despite the increased incidence of ventricular fibrillation associated with multifocal ventricular premature beats following myocardial infarction, no data show improved outcomes as a result of their suppression by antiarrhythmic drugs.

Figure 21.16 Rhythm strip showing ventricular premature beat (see text for details). S, sinus beat; VPB, ventricular premature beat; x, normal RR interval.

Accelerated Ventricular Rhythm

Accelerated ventricular rhythm (accelerated idioventricular rhythm, or slow ventricular tachycardia) is a broad-complex arrhythmia that arises due to enhanced automaticity within the bundle branches or fascicles of the Purkinje system (Fig. 21-17). It is associated with a heart rate of between 50 and 100 beats per minute and is usually self-terminating. Accelerated ventricular rhythm most commonly occurs in the early period following myocardial infarction or as a consequence of digoxin toxicity. There is AV dissociation, with the atrial and ventricular rates typically being relatively similar.

Figure 21.17 Rhythm strip showing accelerated ventricular rhythm. The heart rate is approximately 90 beats per minute. No P waves are visible, and the QRS width is 0.12 seconds.

ELECTRICAL CARDIOVERSION

Cardioversion is a safe and highly effective therapy. Recently, the development of machines that deliver biphasic shocks has further increased the efficacy of cardioversion, compared with traditional direct-current monophasic machines. There are different types of biphasic waveforms, but it is not clear that any one of them is superior in terms of efficacy or reduced myocardial damage.

Tachycardias with a macro reentrant circuit, including atrial flutter, paroxysmal supraventricular tachycardias, and most instances of ventricular tachycardia, can be converted with low energies (20 to 50 joules monophasic, 20 to 40 joules biphasic, but see subsequent recommendations). Atrial fibrillation requires higher energies (100 to 360 joules monophasic, 100 to 200 joules biphasic). Arrhythmias such as accelerated junctional rhythm and some ectopic atrial tachycardias are not amenable to electrical cardioversion.

Despite the safety of cardioversion, certain precautions must be taken. Delivery of an electrical shock during the repolarization period (on the T wave of the ECG) can induce ventricular fibrillation. Thus, for organized rhythms (the various supraventricular tachycardias and monomorphic ventricular tachycardia), synchronized shocks, in which the shock is timed to the R wave of the ECG, should be delivered. For disorganized rhythms (ventricular fibrillation and polymorphic ventricular tachycardia), attempts to deliver a synchronized shock will be frustrated if no R wave can be identified. Thus, nonsynchronized shocks should be used for defibrillating disorganized rhythms.

With rapid heart rates, low-energy synchronized shocks may cause ventricular fibrillation. Shocking with high energies (above the upper limit of vulnerability ²⁹) eliminates this risk. Thus, a general recommendation is that all elective cardioversions use energies of 200 joules monophasic or biphasic. The technique for cardioversion is described in Chapter 40.

Following a shock, a surge of vagal tone may occur, resulting in a short period of asystole. Each shock also transiently impairs fast sodium channel conduction, so that multiple successive shocks may cause a degree of myocardial stunning or even induce temporary electromechanical dissociation (pulseless electrical activity) if myocardial function is impaired.

If sinus rhythm is transiently restored but quickly degenerates back into the original arrhythmia, it denotes an unstable substrate. If this happens with two successive shocks, further energy delivery is pointless until either an appropriate antiarrhythmic drug is given or the hemodynamic state improves. Three unsuccessful shocks with appropriate energies (with not even transient reversion) are an indication to stop electrical cardioversion and reevaluate therapy.

ATRIOVENTRICULAR BLOCK

AV block is divided into first, second, and third (or “complete”) degrees.

First-Degree AV Block

First-degree AV block is diagnosed when the PR interval exceeds 0.2 seconds (5 small squares on the ECG; Fig. 21-18). It occurs as a consequence of disease in the AV node and is common in older patients. In the context of an aortic root abscess, first-degree AV block with a rapidly lengthening PR interval (over days) may signal the imminent development of complete AV block. In general, no treatment is required for first-degree AV block unless prolongation of the PR interval is extreme (>400 ms) or rapidly evolving, in which case pacing is indicated. Prophylactic antiarrhythmic drug therapy is best avoided in patients with marked first-degree AV block.

Figure 21.18 Rhythm strip showing first-degree heart block. The PR interval is extremely prolonged at 360 ms (9 small squares).

Second-Degree AV Block

Wenckebach Second-Degree Block (Mobitz Type I)

Wenckebach AV block is reasonably common in older patients with heart disease. There is progressive prolongation of the PR interval with eventual failure to conduct a P wave, followed by resumption of conduction (Fig. 21-19). The PR interval after the blocked beat is always shorter than the PR interval immediately before it. Wenckebach block usually reflects AV nodal disease but is not an automatic indication for pacing. When second-degree block has appeared for the first time following cardiac surgery and is associated with bradycardia, permanent pacing should be considered.

Figure 21.19 Rhythm strip showing Wenckebach (Mobitz type I) second-degree heart block. The P waves occur at regular intervals at a rate of approximately 60 beats per minute. The PR interval progressively lengthens over the first three beats; after the fourth P wave, no conduction occurs. This is indicated by the absence of the QRS complex. After the fifth P wave, the PR interval is normal.

Mobitz Type II Block

Mobitz type II block is uncommon and is usually associated with newly acquired bundle branch block. Mobitz type II block manifests as abrupt failure of P wave conduction (dropped beats) without PR interval shortening of the subsequent conducted beat. It represents disease in the His-Purkinje system and is an indication for permanent pacing.

2:1 AV Block and High-Grade AV Block

2:1 AV block describes bradycardia in which every second P wave fails to conduct to the ventricles (Fig. 21-20). Block may be either at the level of the AV node or in the His-Purkinje system. With the latter, there is nearly always associated bundle branch block. Persistent block requires permanent pacing to restore normal heart rate. Transient 2:1 AV block may occur during sleep or because of a surge of vagal tone (e.g., during vomiting).

Figure 21.20 Rhythm strip showing 2:1 second-degree heart block. P waves (arrows); every second P wave is conducted.

Complete, or Third-Degree, AV Block

With complete heart block there is complete failure of conduction through the AV node and dissociation between the atria and the ventricles (Fig. 21-21). P waves can be seen marching through the QRS complex. The width and frequency (i.e., heart rate) of the QRS complex depends upon the level within the conducting system at which the escape rhythm originates. If the escape rhythm is near the AV node, the heart rate is typically about 50 beats per minute and the QRS relatively normal. An escape rhythm within the His bundle results in a heart rate of 30 to 40 beats per minute, and the QRS complex may be normal or widened. An escape rhythm that originates below the His bundle is very wide and very slow. Intermittent cannon A waves may be seen on the CVP trace.

Figure 21.21 Rhythm strip showing third-degree (complete) heart block. Marked ST segment elevation is seen, suggesting an inferior (lead II) myocardial infarction. The P waves (arrows) are completely dissociated from the QRS complexes.

The treatment for complete heart block is pacing. In the absence of pacing wires, a bolus dose of atropine (0.6 mg, repeated up to 0.04 mg/kg) combined with an infusion of epinephrine may be used; isoproterenol may also be used but commonly causes hypotension (see Chapter 3). Transcutaneous pacing (see subsequent material) may be necessary until more definitive treatment (e.g., transvenous or permanent pacing) is available.

Bundle Branch Block and Fascicular Block

The His bundle gives rise to the left and right bundle branches, with the left bundle further dividing into anterior and posterior fascicles (see Chapter 1). Thus, there are three terminal branches (or fascicles), one on the right and two on the left. Block of any one branch is called unifascicular block and is not a risk factor for complete AV block. Block of any two branches—either complete LBBB or RBBB with associated block of the anterior (common) or posterior (rare) fascicle—is termed bifascicular block. Block of the anterior fascicle is termed left anterior hemiblock, and block of the posterior fascicle is termed left posterior hemiblock. ECGs showing LBB and RBBB are shown in Figure 21-22 and Figure 21-23, respectively; the ECG criteria for bundle branch and fascicular block are listed in Table 21-1.

Figure 21.22 12-lead ECG showing left bundle branch block. The QRS duration is approximately 140 ms (3.5 small squares). A slurred S wave (QS wave) is seen in V1. No Q wave and a monophasic R wave are seen in V6. No Q wave is seen in lead I.

Figure 21.23 12-lead ECG showing right bundle branch block. The QRS duration is 120 to 140 ms. A late secondary R wave (rSR) is seen in V1, and a late slurred S wave is seen in both V6 and lead I.

Table 21-1 ECG Criteria for Bundle Branch and Fascicular Block

Unifascicular Block
Anterior fascicle of the left bundle branch: left anterior hemiblock	Normal QRS duration
	Left axis deviation
	Small Q in leads I and aVL
	Small R in leads II, III, and aVF
Posterior fascicle of the left bundle branch: left posterior hemiblock	Normal QRS duration
	Right axis deviation
	Small R in leads I and aVL
	Small Q in leads II, III, and aVF
	No evidence of right ventricular hypertrophy (see Chapter 8 and Fig 24.1)
Right bundle branch block (see Fig. 21-23)	QRS	>0.12 sec
	Lead V1	Late secondary R wave (rSR)
	Lead V6	Late slurred S wave
	Lead I	Late slurred S wave
Bifascicular Block
Left bundle branch block (see Fig. 21-22)	QRS	>0.12 sec
	Lead V1	Wide slurred S wave or QS complex
	Lead V6	No Q wave
		Broad, monophasic or notched R wave
	Lead 1	Broad, monophasic or notched R wave
Right bundle branch block with left anterior hemiblock (common)	QRS >0.12 sec
	Left axis deviation
	Appearances of right bundle branch block in lead V₁
	Initial R wave and prominent S wave in leads II, III, and aVF
Right bundle branch block with left posterior hemiblock (uncommon)	QRS > 0.12 sec
	Appearances of right bundle branch block in lead V₁
	Right axis deviation
	Initial R wave and prominent S wave in leads I and aVL

Bundle branch block can occur intermittently, which usually takes place during a time period just before permanent bundle branch block is established. Tachycardia-dependent block can also occur because as the heart rate increases, one of the bundles may still be refractory when the next impulse arrives. When RBBB is associated with left axis deviation, bifascicular block is nearly always the reason. In contrast, the combination of RBBB and right axis deviation is more commonly indicative of right ventricular hypertrophy rather than involvement of the posterior fascicle of the left bundle.

Bifascicular block alone is not an indication for prophylactic pacing; it carries a 1% to 2% annual incidence of progression to complete AV block. Block in all three branches constitutes trifascicular block and results in complete AV block. When bifascicular block is associated with first-degree AV block, the risk for subsequent complete AV block is increased.

LBBB is usually indicative of structural heart disease, particularly coronary artery disease, whereas RBBB may be a normal variant. Occasionally, RBBB develops during insertion of a pulmonary artery catheter. Thus, LBBB is a relative contraindication to pulmonary artery catheterization (see Chapter 8).

Atrioventricular Block Following Cardiac Surgery

Components of the conducting system may be disrupted during cardiac surgery, causing bradycardia or AV block. The SA node may be damaged during placement of superior vena cava cannulas, during right atrial incision (e.g., to gain access to the mitral valve via the interatrial septum), and during cardiac transplantation. The AV node may be damaged during surgery involving the tricuspid valve (especially tricuspid valve replacement) and during repair of a primum atrial septal defect. The His bundle may be damaged during aortic valve or left ventricular outflow tract surgery or as consequence of an aortic root abscess. Transient AV or SA node block is common following cardiac surgery because of the effects of cardioplegia. Sinus bradycardia and first-degree AV block are usual findings, but temporary complete heart block also occurs. For these reasons temporary atrial and ventricular epicardial pacing wires are commonly placed on the heart at the time of surgery.

There is an important issue to consider when complete heart block occurs following cardiac surgery: when should a permanent pacemaker be placed? Indications for permanent pacing in this situation are:

• When preexisting bifascicular block and complete postoperative block persist more than 24 hours after return to the ICU

• When complete AV block persists for more than 72 hours, regardless of the preoperative ECG appearance.

• When AV block of 24 to 72 hours duration occurs, individual review is indicated that takes into account the patient’s general condition, the presence of baseline conduction disease, concomitant drugs that have conduction effects, and the severity of associated heart disease.

Sick Sinus Syndrome

Sick sinus syndrome is a term that includes inappropriate sinus bradycardia, sinus pauses, sinus node exit block, and bradycardia-tachycardia syndrome. In sinus node exit block, pauses of the same duration, or multiples of the sinus rate, without preceding P waves are seen. In bradycardia-tachycardia syndrome, bouts of tachycardia (atrial fibrillation, atrial flutter, AV node reentry) are interspersed with bradycardia. Sick sinus syndrome is the most common indication for permanent pacing.

Sinus pauses are common in the elderly, particularly during sleep. Sick sinus syndrome may become apparent for the first time following cardiac surgery. Initial treatment in the ICU involves a combination of pacing and antiarrhythmic drugs, along with referral to an appropriate cardiologist. Long-term treatment includes:

• A conservative approach in which antiarrhythmic drugs are temporarily withdrawn and the patient observed; the risk for fatal asystole is much lower in sick sinus syndrome than in AV block.

• Implantation of a permanent pacemaker with or without long-term antiarrhythmic therapy. In a third of patients with sick sinus syndrome, AV conduction is also abnormal, and a dual-chamber pacemaker is required. Otherwise, atrial pacing is sufficient if there is no bundle branch block, and the Wenckebach point (the atrial rate at which Wenckebach AV block occurs) is above 120 beats per minute.

TEMPORARY EPICARDIAL PACING

Commonly, following cardiac surgery, four pacing wires are sutured to the heart, two each to the right atrium and right ventricle. The leads are then attached to the positive and negative terminals of the atrial and ventricular outputs of the temporary pacing unit. However, only one lead, conventionally connected to the negative terminal of the pacemaker, must be actually attached to the chamber being paced. The positive terminal may be attached to any conducting surface in the body such as a subcutaneous pacing skin suture or another cardiac chamber. Although it may be possible to pace using the positive terminal attached to the heart (anodal pacing), the thresholds will be higher (particularly with a skin electrode) and capture is less likely. Indications for conversion to a permanent pacemaker are described earlier.

Pacing Codes

A five-letter system is used to describe pacemaker function. The first letter refers to the chamber paced, the second letter refers to the chamber sensed, and the third letter refers to the response to sensing. The fourth and fifth letters refer to programmability and antitachycardia functions respectively, but are not relevant to temporary pacing in the ICU. The chambers paced include the atria (A), the ventricle (V), or both (dual, D). Sensing refers to the detection of electrical activity in the heart by the pacemaker. Sensing may be of A, V, D, or neither chamber (O). If sensing is active, the response may be inhibition of the pacemaker (I), triggering of the pacemaker (T), or both inhibition and triggering (D). In inhibition, a pacing impulse is generated only if the heart rate falls below the set rate of the pacemaker. If sensing and inhibition are active, the pacemaker is said to be in a demand mode. If sensing is not active, the pacemaker is said to be in asynchronous mode. Triggering occurs when both the atrium and the ventricle are sensed and, if no QRS complex follows the P wave within a set time period (the set AV interval), the ventricle is paced.

Commonly Used Pacing Modes

Atrial Pacing

Atrial pacing (AOO and AAI) maintains the benefits of atrial systole but requires normal AV conduction. In the early postoperative period, AV conduction may be generally intact but there may be marked PR prolongation. In this situation, intermittent second- and third-degree AV block can occur, and atrial pacing is not recommended.

In AOO mode, the atrium is paced at a set rate regardless of atrial activity; this is asynchronous atrial pacing. In AAI mode, the atrium is sensed and the pacemaker is inhibited if the native atrial rate rises above the pacemaker set rate; this is demand atrial pacing. In general AAI is preferred to AOO because with AOO, pacing can occur during atrial repolarization, possibly inducing atrial fibrillation. However, AAI mode can potentially be inhibited by extraneous electrical activity, particularly surgical diathermy. Atrial pacing is used for sinus bradycardia and as prophylaxis against atrial fibrillation.

Ventricular Pacing

Ventricular pacing (VOO and VVI) does not depend on AV conduction but there is loss of atrial systole. In situations in which atrial systole is critical (e.g., patients with diastolic dysfunction), ventricular pacing, particularly at high heart rates, can result in profound hypotension.

VOO pacing is asynchronous pacing of the ventricle, irrespective of native ventricular activity. With VVI the ventricle is sensed and the pacemaker is inhibited if the native ventricular rate rises above the pacemaker set rate; this is demand ventricular pacing.

Wherever possible, VOO pacing should be avoided, as there is the possibility of inducing ventricular fibrillation as the result of inadvertent pacing on a native or ectopic T wave (the R-on-T phenomenon). VVI pacing is appropriate for patients in atrial fibrillation when they have slow rates or when there is no capture from atrial wires.

Dual-Chamber Pacing

Dual-chamber pacing (DOO, DDD) has the advantages of maintaining atrial systole and not requiring intact AV function. Despite this, if AV function is normal, atrial pacing, which utilizes native AV conduction, is more hemodynamically efficient than dual-chamber epicardial pacing.

DOO is asynchronous dual-chamber pacing; like VOO, it carries the risk for inadvertent ventricular fibrillation. In DDD mode, both chambers are sensed and paced. The pacemaker senses native atrial activity, and if the rate is greater than the set pacemaker rate, the pacemaker is inhibited. If the atrial rate is less than the set rate, the atrium is paced. If AV conduction is intact (i.e., if it takes less time than the AV interval that is set on the pacemaker), ventricular pacing is inhibited. If AV conduction is absent or is delayed for longer than the set AV interval, ventricular pacing is triggered. In DDD mode, native sinus rhythm above the set rate, in the presence of normal AV conduction, completely inhibits the pacemaker.

A potential problem that may occur with any pacing mode that uses triggering is rapid pacing of the ventricle in the presence of fast atrial activity (e.g., atrial flutter). Thus, an upper rate must be set, above which the ventricular pacing will not track rapid atrial activity. This is usually set at 120 beats per minute.

Pacemaker Settings in Cardiac Surgery Patients

When setting up temporary epicardial pacing, the following parameters must be set.

Mode

Only three pacing modes are commonly used in the ICU:

1. AAI. This mode is used for sinus bradycardia with a relatively normal PR interval and for prophylaxis of atrial fibrillation.

2. VVI. This mode is used for slow atrial fibrillation or failure of atrial capture.

3. DDD. This mode is the default “workhorse” option that is most commonly used in cardiac surgery patients. It is used for bradycardia in the presence of AV block. Although first-degree AV block potentially can be treated with AAI, DDD is preferred because intermittent second- or third-degree AV block is common in the early postoperative period.

Rate

For most patients, a rate of 70 to 90 beats per minute is appropriate to optimize cardiac output and blood pressure. In patients with impaired diastolic function, slightly slower rates may be appropriate. Rates above 90 beats per minute increase oxygen demand and carry a possible risk for myocardial ischemia. In convalescent patients in whom there is persisting concern over the integrity of AV conduction or bradycardia, a backup rate of 60 (usually in DDD mode) is desirable so as to avoid unnecessary overdrive suppression and possible syncope if AV conduction fails. For DDD pacing, an upper rate of 120 to 130 beats per minute is adequate.

Output

Outputs range from 0 to 20 milliamps (mA). This output is delivered over a finite period of time—the pulse width—which is usually set at 1.5 ms. The capture threshold is the lowest current setting at which pacing occurs. The output should be set to 5 mA above the capture threshold. Thresholds tend to rise progressively after the first 1 to 2 days and must be measured at least daily. Thresholds increase, sometimes steeply, with hypoxemia, acidemia, ischemia, and hypothermia.

AV Interval

For dual-chamber pacing, the AV interval must be set. The normal AV delay (i.e., the PR interval) at rest is 120 to 140 ms, with an upper limit of 180 ms and shortening with exercise. When using DDD mode in routine cardiac surgery patients, the AV interval should be set to be relatively long (160 to 180 ms). In this way, as normal AV conduction recovers and the native PR interval falls below the set AV interval, ventricular pacing is spontaneously inhibited. Because cardiac function is optimal at an AV interval shorter than 160 to 180 ms, in very sick patients, the AV interval should be set much lower (e.g., about 120 ms), and periodic checks of native AV conduction should be made.

Sensitivity

The word sensitivity refers to the electrical signal (in mV) that is interpreted by the pacemaker as being atrial or ventricular activity. In asynchronous pacing, sensitivity is set to infinity (in practice, 10 mV) and all native electrical activity is “ignored.” As the sensitivity is gradually reduced to 0 mV, asynchronous pacing becomes demand pacing. Standard presets (e.g., 0.5 mV for atrial sensitivity and 2.5 mV for ventricular sensitivity in DDD pacing) are built into the default settings for the various demand-pacing modes and only rarely need altering.

When sensing is occurring normally, the sensing light will flash on the temporary pacing unit. If the pacemaker is being inappropriately inhibited by extraneous electrical activity (oversensitive) or the pacemaker is failing to be inhibited by native atrial or ventricular activity (undersensitive), it may be necessary to determine the sensing thresholds. To do this, the sensitivity value should set at 0 mV and the pacing rate set below the heart rate. Output should be minimal to prevent capture while this is done. The sensitivity value is then gradually increased (pacemaker becoming increasingly less sensitive) until the sensing light no longer flashes regularly. This is the sensing threshold. The sensitivity value should be decreased at least 1 mV below this value. When the patient is highly pacemaker dependent with no escape rhythm, ventricular sensing thresold should not be tested but the sensitivity value nominally set at 10 mV (asynchronous pacing).

Testing for the Presence of an Escape Rhythm

When pacing is begun for AV block, it is essential to check periodically for resumption of conduction or the presence of an escape pacemaker. Because of overdrive suppression, if temporary pacing is stopped abruptly without prior rate reduction, asystole commonly occurs, and it may be incorrectly interpreted as pacemaker dependence. To test for an escape rhythm, the pacing rate should be decreased from a rate of 60 beats per minute by 10 beats per minute every 30 seconds to about 30 beats per minute to look for the emergence of normal rhythm or an escape rhythm. If there is still only paced rhythm, output can be briefly inhibited for 3 to 4 seconds to see if there is intrinsic activity; then pacing at a normal rate is resumed.

Troubleshooting Temporary Epicardial Pacing

The two problems that commonly complicate temporary epicardial pacing are failure to pace (capture) and competition between the pacemaker and the native rhythm.

Failure to Capture

Failure to capture may arise because of a problem with the epicardial wires, which may become disconnected from the heart or, over time, become fibrosed and ineffective. In one study, loss of capture occurred in more than 50% of patients by postoperative day 5.³⁰ In the first instance, the outputs of the pacemaker should be increased to their maximum values (20 to 25 mA). Reversing the polarity of the leads may help, or connecting each epicardial lead in turn to the negative pole of the pacemaker and using a subcutaneous pacing suture for the positive lead. Correction of electrolyte and acid-base disturbances may help. In DDD or AAI mode, failure to capture sometimes occurs because the pacemaker is over sensitive (sensitivity value too low) and inappropriate inhibition occurs; this problem can be rectified by increasing the sensitivity (see earlier discussion). Failure to capture the atrium in DDD mode is occasionally improved by changing to DVI mode, in which the atrial component is asynchronous. If pacing is essential and these interventions are ineffective, placement of a permanent pacemaker or a temporary transvenous pacing wire may be necessary.

Competition Between Patient and Pacemaker

Competition between the pacemaker and the patient may manifest as an irregular heart rate and a chaotic ECG showing both paced beats and native beats. If the patient is being paced in an asynchronous mode, the pacemaker should either be turned off or changed to a demand mode. If the patient is already in a demand mode, the sensitivity value should be (see previous information) reduced further toward zero.

Other Types of Pacing

Ventricular Pacing

In addition to epicardial pacing, temporary ventricular pacing may be achieved via a transvenous pacing wire, a pacing pulmonary artery catheter or, in emergency situations, with transcutaneous pacing. When pacing occurs via a transvenous wire or a pulmonary artery catheter, the pacing wire is in contact with the right ventricular endocardial surface. Compared to epicardial pacing, it is usual that only relatively low outputs (<5 mA) are required for capture. VVI is the usual mode. Transcutaneous pacing via surface pads is used in emergency situations, usually for complete heart block. The default mode is VOO. Because of the high impedance of the chest cavity, high outputs are required and capture can be difficult. It is best to start at about 200 mA and then reduce the output once capture has been achieved. Transcutaneous pacing is very uncomfortable for patients who are awake, so sedation should be given.

Esophageal Pacing

It is possible to capture the atria, but virtually never the ventricles, by pacing at high (≥20 mA) energies and wide pulse widths (6 to 9 ms) from within the esophagus, which is anatomically behind the atria. A bipolar permanent pacing lead is placed in the distal esophagus at a distance of about 45 cm from the lips. The output is set to a maximum level and the lead is very slowly withdrawn until atrial capture occurs, usually at a distance of about 35 cm from the lips. Occasionally, when only ventricular epicardial pacing wires are present, an esophageal (“atrial”) lead may be used in conjunction with the ventricular lead to facilitate DDD pacing.

Management of Permanent Pacemakers

In patients with permanent pacemakers undergoing cardiac surgery, it is usual to reprogram the pacemaker into demand mode at a low rate (e.g., DDD or VVI at 60 beats per minute) prior to returning to the ICU. Postoperative cardiac pacing is then managed in the usual fashion with temporary epicardial pacing. Care must be taken to avoid pacemaker “cross-talk,” in which the output of one pacemaker is sensed by the other as native atrial or ventricular activity, even though the first pacemaker may not be capturing the atria or ventricle. No effective pacing will then occur. On or about the third postoperative day, once the patient’s rhythm has stabilized and the risk for postoperative atrial fibrillation is lower, the internal pacemaker may be reprogrammed to the preoperative rate and mode, and the temporary epicardial pacing can be discontinued.

For most, but not all, internal pacemakers, placement of a magnet over the pacing box (usually in the subclavicular space) converts the pacemaker into an asynchronous mode (e.g., DOO at 80 beats per minute). This technique should be used only in emergency situations.

Implantable Cardioverter-Defibrillators

The superiority of implantable cardioverter-defibrillators over antiarrhythmic drug therapy for survivors of cardiac arrest is well established.³¹ Prophylactic insertion of implantable cardioverter-defibrillators also confers a survival benefit in patients with coronary artery disease and severe left ventricular impairment or congestive cardiac failure.³² Implantation is a low-risk procedure (mortality rate <1%) and patients have a good quality of life. Therapy is flexible, with initial antitachycardia pacing successful in upwards of 80% of cases of ventricular tachycardia, followed by low-energy cardioversion and high-energy defibrillation shocks. Dual-chamber devices that allow full flexibility in programming are especially useful when there is coexistent atrial flutter or fibrillation. A small subset of cardiac surgery patients with impaired ventricular function and persistent arrhythmias are candidates for early implantation.

In a patient who has an implantable cardioverter-defibrillator in situ, it is essential to inactivate antitachycardia therapies immediately prior to surgery because they may be activated by diathermy; they should be turned back on again when the patient leaves the operating room, and a full routine test should be made to exclude any inadvertent malfunction.