American College of Cardiology

GREGORATOS ET AL., ACC/AHA/NASPE 2002 Guideline Update for Implantation of Cardiac Pacemakers and Antiarrhythmia Devices
http://www.acc.org/clinical/guidelines/pacemaker/index.htm; 2002

ACC/AHA/NSAPE 2002 Guideline Update for Implantation of Cardiac Pacemakers and Antiarrythmia Devices—Full Text

A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (ACC/AHA/NASPE Committee on Implantation)

This is a Guideline Update of the 1997 Implantation of Cardiac Pacemakers and Antiarrythmia Devices Guidelines. To highlight the changes, deleted text is indicated by strikeout, and revised text is presented in brown. A clean version of the document, with changes fully incorporated, is available for download and print.

II. Indications for Implantable Cardioverter-Defibrillator Therapy

A. Background

ICDs were originally developed and have been most frequently used forto prevent sudden cardiac death in patients who have experienced life-threatening ventricular arrhythmias such as sustained VT or VF (202-205). Epidemiological studies report high rates of recurrence of these life-threatening arrhythmias (30% to 50% in 2 years) during follow-up. Early observational reports documented efficacy in reversion of sustained VT and VF (103-105,202,203,205-215,405-407). have now been supplemented by Large prospective, and sometimes randomized single-center and multicenter studies with long-term outcome data have established the effect of the ICD on long-term outcomes (204,216-221).

Enrollment in these trials has included patients with coronary and noncoronary heart diseases with a wide range of ventricular function and coexisting disorders.

These studies uniformly document sudden cardiac death recurrence rates that average 1% to 2% annually after device implantation in these populations. Simultaneously, rapid technological evolution of ICD systems has occurred. The ICD has evolved from a short-lived nonprogrammable device requiring a thoracotomy for lead insertion into a multiprogrammable antiarrhythmia device implanted almost exclusively without thoracotomy, now capable of treating bradycardia, VT, VF, and atrial tachycardia (222-224,408,409). Clinical studiesregistries have recorded major improvements in implant risk, system longevity, symptoms associated with arrhythmia recurrences, quality of life,and diagnosis and management of inappropriate device therapy (103,216-218,225-229).

Implantation, follow-up, and replacement of these devices is a complex process requiring familiarity with device capabilities, adequate case volume, continuing education, and skill in the management of ventricular arrhythmias, therefore mandating involvement of a trained electrophysiologist (230,410,411) to provide an optimal personnel team for patient safety and device management.

A substantial new body of information has emerged regarding the clinical outcome of patients with VT or VF treated with currently available antiarrhythmic therapies. There are currently three major therapeutic options to reduce or prevent VT or VF in patients at risk for these arrhythmias. These are:

1. Antiarrhythmic drug therapy. selected by electrophysiologic study or ambulatory monitoring or prescribed empirically.
2. For VT, ablative techniques applied at cardiac surgery or percutaneously using catheter techniques.
3. Implantation of a cardioverter-defibrillator device system.

A combination of ICD therapy with drugs or ablation is also frequently used. Currently, the largest clinical experience is with combined antiarrhythmic drug and ICD therapy.

B. Clinical Efficacy of ICD Therapy

ICD devices have been extensively evaluated in prospective clinical trials and clinical experience now exceeds device registries 100,000 implants worldwide since the inception of this therapy (103-105,202-221). ICDs have been clearly documented to revertSingle-chamber nonthoracotomy systems can be implanted with a procedural mortality rate of 0.5% (217) to 0.8% (216). The ICD has been shown to terminate VF successfully in 98.8% (217) and 98.6% (216) of episodes. They effectively terminate sustained ventricular tachyarrhythmias, including by pace or shock termination of sustained VT, and shock reversion termination of VF. Large series have shown that single chamber nonthoracotomy systems can be implanted with an average a procedural mortality of 0.5% (217) to 0.8% (216). The ICD has been shown to terminate VF successfully in 98.8% (217) and 98.6% (216) of episodes. VT has been converted with antitachycardia pacing in 89.4% (216) to 91.2% (217) of episodes, with further successful conversions (98%) using shock therapy. Inappropriate therapy, typically for atrial fibrillation with a rapid ventricular response, has been noted in 5% to 11% of patients.

Early retrospective reports showed significant improvements in survival with the defibrillator (205,208,231). The study design tended to overestimate benefits by using device therapies (antitachycardia pacing and shocks) as surrogate mortality events. In a large body of subsequent experience, the sudden death rate reported in virtually all series ranges from 1% to 2% per year with a cumulative incidence of #10% at 5 years (105,205-207,210,211,213,215-217,232). Higher sudden death rates have been reported in patients with severe LV dysfunction (233,234). Dilution of the survival benefit conferred by sudden death reduction in ICD patients by non-arrhythmic mortality and its impact on overall survival is patient population-dependent (232,233).

Device therapy delivery cannot be used as a surrogate mortality end point because arrhythmias other than VT/VF can activate the device, and recurrent VT is not invariably fatal. Symptomatic ICD activations alone underestimate antiarrhythmic benefits of ICD therapy. More recently, firmer estimates of benefits from ICD therapy using devices with event memory capabilities have become possible in the absence of placebo-controlled studies (225,227,235,236). In these studies ICD patients had successful reversion (>98%) of VT with circulatory collapse or VF, with a significant projected survival benefit compared with untreated populations (235). This benefit is incremental and continues to increase over longer periods (3 to 4 years). A similar benefit exists in patients with sustained VT (236).

There has been controversy about the appropriate end point for evaluation of ICD efficacy. Many studies have used sudden death, but classification of the cause of death is often difficult and imprecise. Consequently, it is now accepted that total mortality is the appropriate primary end point for judging ICD efficacy (237). Rates of sudden death and ICD discharges provide useful information, but they should be considered as secondary end points. Total mortality varies significantly between reports due to differences in the disease status of the population under study and LV function. The presence of concomitant cardiac disease is a major determinant of survival (233,238,239). Survival of ICD recipients is influenced by LV function. Patients with LV ejection fractions <30% have reduced survival compared with those with higher ejection fractions at 3 years (233,234). However, both populations appear to derive a significant survival benefit (221). Patients with reduced LV systolic function appear to benefit the most from ICD therapy (412,413). A retrospective analysis indicated that patients thought to have had a reversible cause for their life-threatening arrest were still at substantial risk of death (414).

C. Alternatives to ICD Therapy

Pharmacological options for guidedantiarrhythmic therapy include drugs in classes I, II, and III. Therapy can be guided by Holter monitoring or serial electrophysiologic testing or given empirically. High arrhythmia recurrence rates and moderate sudden death rates are observed with Class I agents (240). By contrast, Class III agents are associated with significantly lower arrhythmia recurrences, sudden death, and total mortality (240-243).

ß-Blocking agents have also demonstrated efficacy in reducing mortality after AMI (244,245). However, their value in a population of patients with sustained ventricular tachyarrhythmias is not well established. Suppression of inducible VT as well as control of spontaneous VT is often not achieved (246). Although the overall survival of cardiac arrest patients treated empirically with beta-blockers and Class I agents may be comparable, patients given Class I agents on serial electrophysiologic testing have a better outcome than those treated with empiric beta-blocker therapy (247). Current data do not support a significant role for monotherapy with beta-blockers in this condition.

In the post–myocardial infarction patient, empiric amiodarone therapy reduces arrhythmic mortality, but benefit with respect to total mortality in such patients with ventricular dysfunction is less clearhas been less consistently demonstrated in individual studies (248-251). In cardiac arrest survivors treated empirically with amiodarone, patients with a reduced ejection fraction (<40%) continue to exhibit high arrhythmia recurrence and sudden death rates (252). Similarly, patients with congestive heart failure may show little to no mortality reduction with empiric amiodarone therapy (253,254). Quantitative overviews of amiodarone use in randomized trials, however, demonstrate a significant reduction in total mortality (415,416).

Long-term maintenance of effective antiarrhythmic drug therapy remains problematic. Discontinuation for drug intolerance is high for Class I agents and sotalol at initiation of therapy and during long-term administration (240). Amiodarone therapy is also frequently discontinued for adverse effects during long-term administration (243).

Ablative therapy has been most often used for patients with sustained monomorphic VT induced at cardiac surgery or electrophysiologic study and mapped to a specific ventricular site(s). Intraoperative ablation is accomplished mechanically or with physical energy sources (cryothermia or laser), whereas catheter-based energy delivery (direct-current shock, radiofrequency, microwaves, or laser) is used during electrophysiologic procedures (255-258). These methods are applicable to a select population of patients with malignant ventricular tachyarrhythmias that have reproducibly inducible monomorphic VT suitable for cardiac mapping. Surgical experience is more extensive and favorable in patients with coronary heart disease than noncoronary disease. Perioperative mortality is now lower and averages <5% in more recent experience, particularly when preoperative LV systolic function is preserved. Intraoperative map-guided ablation—now performed infrequently—is associated with low arrhythmia recurrence (less than 10% at 2 years) and minimal sudden death rates (256-258) during long-term follow-up, making it an important therapeutic alternative in this subgroup. among patients with preserved LV function.

Catheter ablation approaches are still in technological evolution (259,260). Hemodynamically stable VT is usually required for mapping, and radiofrequency energy is currently used for ablation (261,262). Procedural complication rates are moderate with modest arrhythmia control (261,262), often in conjunction with previously ineffective drug therapy in patients with coronary artery disease. Higher efficacy rates are observed in patients with right ventricular outflow tract tachycardia, idiopathic left septal VT, and bundle-branch reentrant VT in whom ablation may be the preferred therapy (263-265). Although acute success rates of 71% have been reported in patients with structural heart disease, recurrence rates of 33% have subsequently occurred (417). Multiple VT morphologies, polymorphic VT, and progressive cardiomyopathy, when present, are less amenable to a favorable result with ablative intervention (255,256). Nonetheless, catheter ablation of recurrent VT associated with frequent ICD therapy may be associated with a marked reduction in the occurrence of shocks and an overall improved quality of life (418). Newer electroanatomic (419) and noncontact (420) mapping technologies may improve the outcomes of such procedures.

D. Comparison of Drug and Device Therapy for Secondary Prevention of Cardiac Arrest and Sustained Ventricular Tachycardia

A significant body of information from prospective, randomized trialsnow exists comparing these two therapeutic optionsICD and drug therapy is now available. Direct comparison of drug and device therapy has been performed in several retrospective nonrandomized reports and fewerin prospective randomized studies. In comparison with concurrently medically treated but nonrandomized populations receiving amiodarone, a significant mortality benefit was noted in the patients with ICDs over the first 3 years of follow- up (206,209). This benefit may dissipate with follow-up beyond 5 years in some reports (209). In similar nonrandomized comparisons in sudden death survivors discharged either on electrophysiologicly guided antiarrhythmic therapy using Class I or III drugs or on an ICD-based regimen, the survival of the ICD patients was superior, both in patients with early or advanced LV dysfunction (210). In such analyses, the use of an ICD in the treatment regimen was the strongest predictor of long-term survival. ICD recipients also show improved survival in such comparisons with patients receiving guided sotalol therapy (266).

Information from randomized trials comparing drug and device therapy also suggests These trials demonstrate survival benefits with ICD therapy in this population compared with electrophysiologically guided drug therapy using Class I agents, propafenone, or sotalol (267,268). The first reported large prospective, randomized study comparing ICD therapy with Class III antiarrhythmic drug therapy, predominantly empiric amiodarone, in survivors of cardiac arrest and hemodynamically unstable VT revealed greater survival with ICD therapy (221). Unadjusted survival estimates for the ICD and drug therapy, respectively, were 89.3% versus 82.3% at 1 year, 81.6% versus 74.7% at 2 years, and 75.4% versus 64.1% at 3 years. Estimated relative risk reduction with ICD therapy was 39% at 1 year and 31% at 3 years. Two other reports of large prospective trials in similar patient groups have shown similar trends (406,407). A pooling of these three studies shows a 27% reduction in total mortality with ICD therapy (421).

Implementation of ICD therapy has been directly compared for safety with antiarrhythmic drug therapy in large systematic trials. Prospective observational data demonstrate a low perioperative mortality (0.4% to 1.8%) for primary nonthoracotomy implants (105,216-218). Similar mortality estimates in large prospective antiarrhythmic drug trials range from 3.2% to 13.0% (221,240,243). However, these populations may not be directly comparable. During long-term therapy, drug discontinuation rates have ranged from 7% to 32%, the lowest being with sotalol in reported data (240). In a large prospective trial, 98% of randomly assigned patients could be maintained on ICD therapy, with 25.4% requiring the addition of drug therapy by 2 years (221). Withdrawal of device therapy is infrequent and rarely exceeds 2% of implants (216-218). The addition of an antiarrhythmic drug in selected patients with ICDs may improve quality of life by reducing arrhythmia recurrences and the need for shock therapy (266,269).

E. Specific Disease States and Secondary Prevention of Cardiac Arrest or Sustained Ventricular Tachycardia

Prior guidelines do not relate the decision to implant an ICD device to the underlying cardiac disease (270). Current information suggests that the underlying disease state may have an important impact on patient prognosis and will influence the decision to implant an ICD earlier or later in the treatment algorithm.

1. Coronary Artery Disease

Patients with coronary artery disease represent the majority of patients receiving devices in most reports. Device implantation is widely accepted as improving the outcome of these patients. Patients with reduced LV function may experience greater benefit with ICD therapy than with drug therapy (208,210,267,412,413). To limit patient risk during defibrillation efficacy testing (270,271), assessment for the presence of active ischemia should precede implementation of device therapy. Furthermore, optimal anti-ischemic therapy (including, where possible, a beta-blocker) will further enhance survival. Measurement of ventricular function is recommended, although poor function is not necessarily a contraindication to device implantation. Abbreviated defibrillation threshold testing, however, may be desirable in patients with elevated pulmonary capillary wedge pressures or severely compromised cardiac output (271).

2. Idiopathic Dilated Cardiomyopathy

Dilated cardiomyopathy is associated with a high mortality within 2 years of diagnosis, with a minority of patients surviving 5 years (272). Approximately one-half of these deaths are sudden and unexpected (273). The combination of poor LV function and frequent episodes of nonsustained VT in these patients is associated with an increased risk of sudden death (274). Moreover, although useful in patients with unlike in ischemic coronary heart disease, the value of electrophysiologic studies in patients with nonischemic cardiomyopathies is limited (275). Furthermore, the efficacy of drug therapy is low in the presence of impaired LV function and difficult to predict on the basis of invasive or noninvasive testing. ICD implantation may be preferred for treatment of patients with VT or VF and this condition. In one large prospective study, this population represented 10% of the study group and showed survival benefits with ICD rather than empiric amiodarone therapy similar to the entire study cohort (221).

3. Long-QT Syndrome

The long-QT syndromes represent a spectrum of electrophysiologic disorders characterized by a propensity for development of malignant ventricular arrhythmias, especially polymorphic VT (239,276-278). Because this is a primary electrical disorder, usually with no evidence of structural heart disease or LV dysfunction, the long-term prognosis is excellent if arrhythmia is controlled. Long-term treatment with beta-blockers, permanent pacing, or left cervicothoracic sympathectomy is frequently effective (277). ICD implantation is recommended for selected patients in whom recurrent syncope, sustained ventricular arrhythmias, or sudden cardiac death occurs despite drug therapy (276). Furthermore, use of the ICD as primary therapy should be considered in certain patients, such as those in whom aborted sudden cardiac death is the initial presentation of the long-QT syndrome, where there is a strong family history of sudden cardiac cardiac death, or when compliance or intolerance to drugs is a concern (276,422).

4. Idiopathic Ventricular Fibrillation

It has been estimated that in 10% of young patients resuscitated from cardiac arrest, the origin of VF is not determined despite extensive evaluation (279,280). Electrophysiologic testing in these patients with “idiopathic VF” usually reveals polymorphic VT or VF that is often suppressible with by Class IA drugs administered during the electrophysiologic testing (279). However, the long-term efficacy of drug therapy remains unknown. Given the guarded prognosis even with effective drug therapy (the annual rate of sudden cardiac death is estimated to be as high as 11%), the limited clinical data available appear to support the use of ICDs in such patients (279-281).

Brugada Syndrome
Individuals with syncope and/or a family history of unexplained sudden cardiac death who have variants of right bundle-branch block QRS morphology and ST-segment elevations in leads V1 through V3 have the Brugada syndrome. Because of the high incidence of VF, ICD therapy should be recommended in such individuals (423). However, ICD therapy is not recommended in individuals with these ECG findings alone, because the specificity of this finding in the absence of the historical details noted above is very low (424).Catheter ablation should be considered before ICD insertion in patients with idiopathic right or left VT (263).

5. Idiopathic Ventricular Tachycardia

Ventricular tachycardia may arise in structurally normal hearts from the right ventricular outflow tract or the LV. These arrhythmias should be treated pharmacologically or with catheter ablation, if amenable, before an ICD is considered for these patients (263).

6. Hypertrophic Cardiomyopathy

Hypertrophic cardiomyopathy should be suspected and is often identified as the cause of sudden death in young people, including trained athletes (239,282). Ventricular tachyarrhythmias are a common mechanism of sudden death in this condition (283). Sudden death may also be the first manifestation of the disease in a previously asymptomatic individual. Criteria to stratify these patients according to risk are not well defined. In contrast with other cardiomyopathies, electrophysiologic testing may be of prognostic value because inducible sustained ventricular arrhythmias appear to be associated with cardiac arrest and syncope in some studies (284)The most prominent characteristics of patients with hypertrophic cardiomyopathy who may be at high risk for experiencing sudden death include the following: 1) prior cardiac arrest or sustained VT; 2) a history of a first-degree relative who has experienced sudden cardiac death; 3) LV hypertrophy with a wall thickness greater than 30 mm (425); 4) syncope, if exertional, repetitive, or in a young patient if no other cause is documented; and 5) nonsustained VT on ECG monitoring if frequent, repetitive, and prolonged. Studies of patients resuscitated from cardiac arrest indicate that many patients will experience another event. Prophylactic pharmacological therapy in the form of betablockers or calcium channel antagonists has frequently been used, but efficacy in sudden death prevention has not been established. Empiric use of amiodarone has been reported to be associated with improved survival in one observational study with historical controls (282). However, prediction of drug efficacy remains difficult and controversial. Sudden death survivors should be considered for ICD therapy in preference to or in conjunction with drug therapy (285). Because these patients are often young, drug compliance is frequently an issue. Long-term protection for these patients may be better afforded by treatment with an ICDOther data support automatic defibrillator implantation as the preferred therapy in high-risk patients with hypertrophic cardiomyopathy to decrease sudden cardiac death, in preference to or in conjunction with drug therapy (285,426).

7. Arrhythmogenic Right Ventricular Dysplasia/ Cardiomyopathy

Arrhythmogenic right ventricular dysplasia can be an important cause of congestive heart failure and ventricular arrhythmias in some patients (286). Drug therapy is often used as primary therapy but is often ineffective. Nonpharmacological options for treatment of significant arrhythmias include catheter ablation of the sites of tachycardia, surgical disarticulation of the right ventricle, and ICDs. In patients with drugrefractory malignant arrhythmias, the ICD provides prophylaxis against syncope due to hemodynamically unstable VT and sudden death (287,288).

8. Syncope With Inducible Sustained Ventricular Tachycardia

Patients with syncope of undetermined etiology in whom clinically relevant VT/VF is induced at electrophysiologic study may be candidates for ICD therapy. In these patients, the induced arrhythmia is presumed to be the cause of syncope (289-291). Cardiovascular mortality averages 20% annually, with a large proportion of it sudden. In some patients, antiarrhythmic treatment is limited by inefficacy, intolerance, or noncompliance. ICD therapy is often used in sustained VT populations with comparable results (292). In patients with hemodynamically significant and symptomatic inducible sustained VT, ICD therapy can be a primary treatment option. The documentation of appropriate ICD therapy of VT and VF from review of event counters and stored electrograms in such patients lends support to use of ICD therapy as a primary treatment option in those who have experienced syncope (414,427).

F. Pediatric Patients

Pediatric experience with ICDs represents less than 1% of all implantations (239,293). Special considerations such as the need for lifelong pharmacological therapy with its associated problems of noncompliance and side effects make the ICD an important treatment option for young patients.

Fewer than 1% of all ICD implantations are performed in pediatric patients (239,293). However, special considerations, such as the need for lifelong pharmacological therapy with its associated problems of noncompliance and side effects, make the ICD an important treatment option for young patients.

Sudden cardiac death is uncommon in childhood but is associated with three principal forms of cardiovascular disease: 1) congenital heart disease, 2) cardiomyopathy, and 3) primary electrical disease (239,294). Patients with pre-existing heart disease are more likely to experience ventricular tachyarrhythmias as the immediate cause of sudden death than are those with normal hearts (295). However, a lower percentage of children undergoing resuscitation survive to hospital discharge compared with adults (296).

Indications for ICD therapy for pediatric patients are similar to those for arrhythmias in adults. However, the data used for risk stratification in adults with coronary artery disease may have less positive predictive value in pediatric patients with a variety of underlying diseases (297). The risk of unexpected sudden death is greater in young patients with diseases such as hypertrophic cardiomyopathy or long-QT syndrome than in adults; therefore, a family history of sudden death may influence the decision to use an ICD in a pediatric patient (277,282). A limited experience with ICDs implanted in young patients with hypertrophic cardiomyopathies or long-QT syndromes after resuscitation has been encouraging (285,293,301,302,351,428).

In patients with congenital heart disease, sudden death has been estimated to occur in 1.5% to 2.5% of patients per decade after repair of tetralogy of Fallot (298). A higher risk has been identified for patients with transposition of the great arteries and aortic stenosis, with most cases presumed to be due to a malignant ventricular arrhythmia associated with ischemia, ventricular dysfunction, or a rapid response to atrial flutter (120,299). An ischemic substrate for arrhythmias leading to sudden cardiac death also exists in congenital coronary anomalies or after Kawasaki disease.

ICD therapy may be preferable to antiarrhythmic drugs in patients with dilated cardiomyopathy or other causes of impaired ventricular function who experience syncope or sustained ventricular arrhythmias because of concern about drug-induced proarrhythmia and myocardial depression. ICDs may also be considered as a bridge to orthotopic heart transplantation in pediatric patients with ventricular arrhythmias whose anomalies are not amenable to surgical correction, particularly given the longer times to donor procurement in younger patients (300). Young patients with hypertrophic cardiomyopathy have a higher annual sudden cardiac death event rate than adults (282,301). A limited experience insuffiwith ICDs implanted in this population after resuscitation has been encouraging (285,293,302).

G. Primary Prevention of Sudden Cardiac Death

1. Coronary Artery Disease

Nonsustained VT in patients with prior MI and LV dysfunction is associated with a 2-year mortality estimated at 30% (303,304). Approximately one-half of this is believed to be arrhythmic in origin. Antiarrhythmic drug therapy has been widely prescribed in patients after MI with and without ventricular arrhythmia, but evidence of improved survival with this approach is not forthcoming. Increased mortality in coronary disease patients with and without nonsustained VT has actually been noted with specific Class I agents (305). Empiric amiodarone therapy has shown inconsistent survival benefit in large prospective randomized trials (250,251), although quantitative overviews (meta-analyses) suggest total mortality may be reduced compared with other medical therapies (241,306). In this population, electrophysiologic testing has identified a subgroup with inducible sustained ventricular tachyarrhythmias that is at high risk for sudden death (307). While arrhythmia-related symptoms and repeated MIs may help identify such patients, asymptomatic persons post-MI may also be at high risk (304,307,308). Survival of patients treated with drugs that suppressed induced arrhythmias improved in comparison with historically untreated or drug-refractory patients (307). In the first prospective randomized trial conducted in such a patient population, improved survival was documented after implantation of ICDs in patients with inducible and nonsuppressible ventricular tachyarrhythmias when compared with conventional drug therapy, including amiodarone (220). Results of another prospective, randomized trial showed reduced mortality with therapy for patients with low ejection fraction, nonsustained VT on Holter monitoring, and inducible sustained ventricular tachyarrhythmias at electrophysiologic study (405). Most of this benefit appeared to be due to ICD placement. The results of a study of 1232 patients after myocardial infarction with an LV ejection fraction of less than or equal to 30% randomized to ICD therapy or not in a 3:2 fashion without the requirement for electrophysiologic screening for inducible ventricular tachyarrhythmia were reported during this publication’s review process (429). At a mean follow-up of 20 months, the mortality rate was 14.2% in the individuals who had ICDs and 19.8% in the conventionally treated group, a 5.6% absolute and 31% relative risk reduction for death. In this study, the survival curves did not begin to separate until 9 months after randomization. Of potential importance, ICD therapy was not implemented until at least 1 month after myocardial infarction and 3 months after coronary artery revascularization surgery. The cost impact, as well as a thorough analysis of the impact of other variables (e.g., risk stratification potential of programmed stimulation performed through the ICDs) has not been fully analyzed and reported. Also of note was an observed increased incidence of new or worsened heart failure in the ICD-treated patients compared with those in the conventional treatment arm. As one editorial suggested, the extent to which ICD therapy actually extends life in given patients is not fully known, and more refined screening techniques even in such patients are needed (430). Although such patients merit consideration of ICD therapy, this approach requires consideration of the patient’s overall health and life expectancy. Additional risk stratification studies are needed to better define which patient subgroups will benefit more or less from ICD therapy than that demonstrated in the abovereferenced population. Further preliminary data presented at the May 2002 Scientific Session of NASPE as a “late-breaking clinical trial” analyzed the effects of ICD therapy in patients stratified on the basis of various noninvasive ECG criteria. A standard ECG QRS duration longer than 0.12 seconds was found to be the strongest predictor of such patients who benefit most from ICD therapy. In the Cox proportional hazard model analysis, individuals with a QRS duration greater than 0.12 seconds had a 63% reduction in mortality relative to conventionally treated patients (p = 0.004). Atrial fibrillation as the baseline rhythm was the only other independent predictor of ICD therapy benefit in such patients. Whether the recommendation to implant ICDs in post–myocardial infarction patients with LV ejection fractions of 30% or less should be limited to individuals with these high-risk variables awaits further clarification.

2. After Coronary Artery Bypass Surgery

Routine ICD insertion does not improve survival in patients with coronary artery disease undergoing bypass surgery who are believed to be at high risk of sudden death based on QRS duration and severe LV dysfunction. In one randomized study, no benefit was noted over placebo (309) in patients with ejection fractions less than 35% and a positive signalaveraged ECG who were undergoing surgical revascularization.

3. As a Bridge to Heart Transplantation

Orthotopic heart transplantation has emerged as an acceptable therapeutic alternative for selected patients with congestive heart failure caused by severe ventricular dysfunction. About 20% of patients requiring transplantation die awaiting a donor organ, with a significant incidence of sudden death. ICDs effectively prevent have been associated with a lower risk of sudden death in these patients (310,311). This benefit is diluted by mortality due to heart failure in some patients (310-312).

4. Other Populations

Other high-risk populations under study for similar benefits include asymptomatic patients, from the standpoint of ventricular tachyarrhythmias, who have impaired LV systolic function and congestive heart failure (431) or idiopathic dilated cardiomyopathy (432), but no recommendations can yet be made with respect to these patients owing to insuffiprecient data. Randomized trials of the ICD are ongoing in these populations. Patients with advanced structural heart disease and syncope of unknown origin may benefit from an ICD even if electrophysiologic evaluation is negative (433).

H. Contraindications and Limitations of to ICD Therapy

ICD therapy is not recommended for the conditions listed below. The first major group, which can be identified by invasive and noninvasive preimplantation testing, includes thosepatients in whom a reversible triggering factor for VT/VF can be definitely identified, such as ventricular tachyarrhythmias in evolving AMI or electrolyte abnormalities. Another population in whom ICD therapy is not routinely recommended is coronary disease patients without inducible or spontaneous VT undergoing routine coronary bypass surgery (309). Similarly, patients with Wolff-Parkinson-White syndrome presenting with VF secondary to atrial fibrillation should undergo catheter or surgical ablation if their accessory pathways are amenable to such treatment.

Patients with terminal illnesses, NYHA class IV drugrefractory congestive heart failure who are not candidates for cardiac transplantation, or with a life expectancy not exceeding 6 months are likely to obtain limited benefit—if any—from ICD therapy. Thus, ICD therapy is discouraged in such individuals. Significant behavioral disorders, including anxiety, device dependence, or social withdrawal, have been described (316,317). A history of psychiatric disorders, including uncontrolled depression and substance abuse that interfere with the meticulous care and follow-up needed by these patients, is a relative contraindication to device therapy.

Patients who have frequent tachyarrhythmias that may trigger shock therapy, such as sustained VT not responsive to antitachycardia pacing or pharmacological therapy, are not suitable candidates for a device because these events would cause frequent device activation and multiple shocks. Alternative therapies, such as combining drugs or ablation with ICD insertion, should be considered.

I. Cost-Effectiveness of ICD Therapy

Several studies have addressed the cost-effectiveness of ICD therapy. The cost-effectiveness ratio compares the total cost of ICD therapy with the total cost of an alternative management strategy such as amiodarone or guided serial drug testing. The overall costs of the ICD have been reduced as the result of nonthoracotomy implantation methods and improvements in ICD reliability and longevity that reduce cost of device replacement and modification. Significant reductions in initial costs have been realized, with newer treatment algorithms eliminating prolonged drug testing (318,319).

The early studies of ICD cost-effectiveness were based on mathematical models and relied on nonrandomized studies to estimate clinical efficacy and cost. These studies found costeffectiveness ratios of $17,000 (320), $18,100 (321), and $29,200 per year of life saved (322). Another model incor-porated costs of nonthoracotomy ICDs and efficacy estimates based on randomized trials and found ICD cost-effectiveness was between $27,300 and $54,000 per life-year gained, corresponding to risk reduction of 40% and 20%, respectively (323).

Several completed and ongoing randomized clinical trials have measured cost as well as clinical outcomes and thus can directly estimate ICD cost-effectiveness. A preliminary analysis of the MADIT (324) trial found the ICD to have a cost-effectiveness ratio of $27,000 per life-year gained. The Multicenter Automatic Defibrillator Implantation Trial (MADIT) found a 54% reduction in total mortality and a cost-effectiveness ratio of $27,000 per life-year added (434). The Canadian Implantable Defibrillator Trial (CIDS), by contrast, found a 20% reduction in total mortality and a costeffectiveness ratio of $139,000 per life-year added (406,435). The cost-effectiveness ratio from the Antiarrhythmics Versus Implantable Defibrillators (AVID) trial was $66,677 per lifeyear added (436). This range of results is primarily due to different estimates of the effectiveness of the ICD in reducing mortality, because all showed similar increases in the cost of care among ICD recipients. When the results of all clinical trials were used in a model that projected the full gain in life expectancy and lifetime costs (323), the cost-effectiveness of the ICD was $31,500 per life-year added,All studies suggest that ICD implantation in appropriately selected patients has a cost-effectiveness ratio comparable to other cardiovascular therapies as well as comparable to widely accepted noncardiac therapies such as renal dialysis ($30,000 to $50,000 per year of life saved). The cost-effectiveness of the ICD is more favorable in patients with high risk of arrhythmic death but low risk of other causes of deathan ejection fraction below 35%. In principle, the device is most cost-effective in patients at high risk of arrhythmic death and at low risk of other causes of death. Cost-effectiveness of the ICD would be improved by lowering the cost of the device itself and further improving its reliability and longevity.

J. Selection of ICD Generators

All ICDs currently marketed in the United States incorporate a number of advanced features, including multiple tachycardia zones, with rate detection criteria and tiered therapy (including low-energy cardioversion and high-energy defibrillation shocks) independently programmable for each zone. Furthermore, all devices incorporate programmable ventricular demand pacing, antitachycardia pacing, and extensive diagnostics, including stored electrograms of rhythms immediately before and after tachycardia detection and therapy. The principal feature distinguishing some ICDs from others is the availability of antitachycardia pacing as a programmable therapy option. The addition of antitachycardia pacing increases the cost of the device by 5% to 10% compared with similar ICDs without this feature. The vast majority of devices are small enough for pectoral new implantations. Larger devices suitable for abdominal implants are available primarily as replacement generators in patients with precient existing lead systems but are being phased out by manufacturers; these larger devices are available at a cost savings of approximately 10% to 25% compared with the smaller devices.

Antitachycardia pacing appears to be a useful feature in the majority of patients receiving ICDs. In one study (325), antitachycardia pacing was activated in 68% of patients receiving ICDs with such a capability, despite the fact that the efficacy of antitachycardia pacing was tested with the device in only 53% of the patients in whom it was activated; in the remainder, antitachycardia pacing algorithms were programmed empirically. In the patients with activated antitachycardia pacing, 96% of all detected episodes of ventricular tachyarrhythmias were terminated with pacing (325). Acceleration of VT by antitachycardia pacing remains a concern, with most series reporting an incidence of antitachycardia pacing acceleration of an episode of VT ranging from 3% to 6% (326). Patients whose only clinical arrhythmia detected before ICD implantation was VF have a lower likelihood of having VT subsequently detected by the ICD than do patients with a prior history of VT (327). However, the incidence of subsequent VT in those with a history of only VF before device implantation is not inconsiderable [18% during 14 months of follow-up in one study (327)], so it is reasonable to select a device with assess activation of antitachycardia pacing even in such patients.

Defibrillators incorporating an atrial lead are now available. Such devices not only provide dual-chamber pacing but also use the pattern of sensed atrial depolarization to distinguish supraventricular from ventricular arrhythmias. A dualchamber pacemaker-ventricular defibrillator device is an appropriate choice for an ICD candidate who has a concomitant need for dual-chamber pacing or a patient with supraventricular tachycardia thought likely to lead to inappropriate ICD therapies.

Early reports have documented that the time required for detection of VF during acute testing at the time of implantation is not impaired with the addition of atrial leads (437,438). However, hemodynamic benefits of dual-chamber pacing in such devices have been documented in few patients followed up for relatively short-term intervals only (439). Furthermore, data on long-term benefits of dual-chamber ICDs are lacking despite their widespread dissemination. One center’s reported experience has been associated with a 2.8% complication rate in their first 95 implants. A worrisome infection rate of 8.8% was observed in patients who had previously implanted single-chamber ICDs upgraded to dual-chamber systems (440). Expected incremental benefits of the use of atrial electrograms together with ventricular electrograms to evaluate stored arrhythmic events have been documented (441). Studies are ongoing to further assess efficacy benefits that dual-chamber ICD therapy may have over single-chamber ICD therapy.

Atrial defibrillation therapy for recurrent AF has been evaluated in stand-alone implantable atrioverters (408) and in conjunction with conventional dual-chamber ICDs (409,442). Although clinically available as complementary features with some dual-chamber ICDs, the indications and role of such therapy are unclear and must await further reports and results of additional studies.

K. ICD Follow-up

All patients with ICDs require periodic and meticulous follow- up to ensure safety and optimal device performance. The goals of ICD follow-up include monitoring of device system function; optimizing performance for maximal clinical effectiveness and system longevity; minimizing complications; anticipating replacement of system components; ensuring timely intervention for clinical problems; patient tracking, education, and support; and maintenance of ICD system records. The need for device surveillance and management should be discussed a priori with patients before insertion of an ICD. Compliance with device follow-up is an important element in evaluating appropriate candidates for device therapy and obtaining the best long-term result. ICD follow-up is best achieved in an organized program analogous to pacemaker follow-up at outpatient clinics (198).

Institutions performing implantation of these devices should also maintain these facilities for inpatient and outpatient use. Such facilities should obtain and maintain implantation and follow-up support devices for all ICDs used at that facility. The facility should be staffed or supported by a fully trained clinical cardiac electrophysiologist (328) who may work in conjunction with trained associated professionals (198,328,329). Access to these services should be available as far as is feasible on both a regularly scheduled and emergent 24-hour-per-day basis. The implantation and/or followup facility should be able to locate and track patients who have received ICDs or who have entered the follow-up program.

1. Elements of ICD Follow-up

The follow-up of an ICD patient must be individualized in accordance with the patient’s clinical status and conducted by a fully trained clinical cardiac electrophysiologist. In general, device programming is initiated at implantation and should be reviewed at predischarge and/or subsequent postoperative electrophysiologic testing. Devices should be followed at 1- to 4-month intervals, depending on the device model and the patient’s clinical status. Manufacturer guidelines for device follow-up vary with individual models and should be available. Transtelephonic follow-up (if available) should always be supplemented by clinic visits at a minimum of 4-month intervals for patient and device evaluation (330,410).

It is often necessary to reprogram the initially selected parameters either in the outpatient clinic or by electrophysiologic testing. When device function or concomitant antiarrhythmic therapy is modified, electrophysiologic testing can be and often is required to evaluate sensing, pacing, or defibrillation functions of the device. Particular attention should be given to review of sensing parameters, programmed defibrillation and pacing therapies, device activation, and event logs. Technical elements requiring review include battery status, lead system parameters, and elective replacement indicators. Intervening evaluation of device function is often necessary. In general, in patients experiencing device activation, with or without therapy, delivery should be evaluated shortly after the event until a regular acceptable pattern of patient symptomatology and tolerance for such events is established and device behavior is deemed reliable, safe, and effective.

After insertion of a device, its performance should be reviewed, limitations on the patient’s specific physical activities established, and registration accomplished. Current policies on driving advise the patient with an ICD to avoid operating a motor vehicle for a minimum of 3 months and preferably 6 months after the last symptomatic arrhythmic event to determine the pattern of recurrent VT/VF (331,332). Interactions with electromagnetic interference sources, impact on employment, and prophylaxis for device infections should be discussed. ICD recipients should be encouraged to carry proper identification and information about their device at all times. Patients receiving these devices can experience transient or sustained emotional disturbances. Education and psychological support before, during, and after ICD insertion are highly desirable and can improve the patient’s quality of life (316,317).

Recommendations for ICD Therapy

Class I

1. Cardiac arrest due to VF or VT not due to a transient or reversible cause. (Level of Evidence: A) (103- 105,202,203,205-211,216,217,219,221,238,260,267, 269,406,407)
2. Spontaneous sustained VT in association with structural heart disease. (Level of Evidence: B) (103- 105,202,203,205-211,216,217,219)
3. Syncope of undetermined origin with clinically relevant, hemodynamically significant sustained VT or VF induced at electrophysiologic study when drug therapy is ineffective, not tolerated, or not preferred. (Level of Evidence: B) (204,213,215,219,227,228, 266,406)
4. Nonsustained VT in patients with coronary disease, prior MI, LV dysfunction, and inducible VF or sustained VT at electrophysiologic study that is not suppressible by a Class I antiarrhythmic drug. (Level of Evidence: BA) (220,308,405)
5. Spontaneous sustained VT in patients without structural heart disease not amenable to other treatments. (Level of Evidence: C)

Class IIa

Patients with left ventricular ejection fraction of less than or equal to 30% at least 1 month post myocardial infarction and 3 months post coronary artery revascularization surgery. (Level of Evidence: B) (429)

Class IIb

1. Cardiac arrest presumed to be due to VF when electrophysiologic testing is precluded by other medical conditions. (Level of Evidence: C) (211,218,267,276)
2. Severe symptoms (e.g., syncope) attributable to ventricular tachyarrhythmias in patients awaiting cardiac transplantation. (Level of Evidence: C) (310,311)
3. Familial or inherited conditions with a high risk for life-threatening ventricular tachyarrhythmias such as long-QT syndrome or hypertrophic cardiomyopathy. (Level of Evidence: B) (8,41,277,282,284,288,300-302)
4. Nonsustained VT with coronary artery disease, prior MI, LV dysfunction, and inducible sustained VT or VF at electrophysiologic study. (Level of Evidence: B) (103,205,212,217,220,307,308)
5. Recurrent syncope of undetermined origin in the presence of ventricular dysfunction and inducible ventricular arrhythmias at electrophysiologic study when other causes of syncope have been excluded. (Level of Evidence: C)
6. Syncope of unexplained origin or family history of unexplained sudden cardiac death in association with typical or atypical right bundle-branch block and STsegment elevations (Brugada syndrome). (Level of Evidence: C) (443,444)
7. Syncope in patients with advanced structural heart disease in whom thorough invasive and noninvasive investigations have failed to define a cause. (Level of Evidence: C)

Class III

1. Syncope of undetermined cause in a patient without inducible ventricular tachyarrhythmias and without structural heart disease. (Level of Evidence: C)
2. Incessant VT or VF. (Level of Evidence: C)
3. VF or VT resulting from arrhythmias amenable to surgical or catheter ablation; for example, atrial arrhythmias associated with the Wolff-Parkinson- White syndrome, right ventricular outflow tract VT, idiopathic left ventricular tachycardia, or fascicular VT. (Level of Evidence: C) (259-263)
4. Ventricular tachyarrhythmias due to a transient or reversible disorder (e.g., AMI, electrolyte imbalance, drugs, or trauma) when correction of the disorder is considered feasible and likely to substantially reduce the risk of recurrent arrhythmia. (Level of Evidence: CB) (414,445,446)
5. Significant psychiatric illnesses that may be aggravated by device implantation or may preclude systematic follow-up. (Level of Evidence: C) (316,317)
6. Terminal illnesses with projected life expectancy less than 6 months. (Level of Evidence: C)
7. Patients with coronary artery disease with LV dysfunction and prolonged QRS duration in the absence of spontaneous or inducible sustained or nonsustained VT who are undergoing coronary bypass surgery. (Level of Evidence: B) (309)
8. NYHA Class IV drug-refractory congestive heart failure in patients who are not candidates for cardiac transplantation. (Level of Evidence: C)

© 2002 by the American College of Cardiology Foundation, American Heart Association, Inc., and North American Society for Pacing and Electrophysiology