The LVEF used in clinical trials assessing the ICD for primary prevention of SCD

< 40% in MUSTT (Multicenter Unsustained Ventricular Tachycardia Trial)

< 30% in MADIT II (Multicenter Automatic Defibrillator Implantation Trial II) (329,332) enrolled 1,232 patients with ICM. No spontaneous or
induced arrhythmia was required for enrollment. All-cause mortality was 20% in the control group and 14.2% in the ICD group (relative risk
31%; p<0.016).

<35% MADIT I (Multicenter Automatic Defibrillator Implantation Trial I) (327)

<35% SCD-HeFT (Sudden Cardiac Death in Heart Failure Trial) (333) included patients with both ICM and NICM, and NYHA Class II or III
congestive heart failure (333). Among the 1,486 patients with ischemic heart disease randomized to either placebo or ICD therapy, the
5-year event rates were 0.432 and 0.359, respectively (HR 0.79; p<0.05). Two recent meta-analyses of these trials have supported the
overall conclusion that ICD therapy in high-risk individuals with coronary artery disease results in a net risk reduction for total mortality of
between 20% and 30% (325,368). The absolute mortality decrease in the medical group was 7.2% after 5 years. The ICD group
experienced a decreased risk of death of 23% compared with the placebo group (HR 0.77, 97.5% CI 0.62 to 0.96), and total mortality in the
medical group was 7.2% per year, with a risk reduction of 23% in the ICD group versus placebo (95% CI 0.62 to 0.96; p<0.007). Relative
risk reduction was comparable for the group with LV dysfunction due to prior MI and the nonischemic group, but absolute mortality was
lower in the nonischemic group. This resulted in a greater number needed to treat per life saved among ischemic patients. There was no
mortality difference between the amiodarone and placebo groups.

CABG-Patch (Coronary Artery Bypass Graft-Patch) trial (328), routine ICD insertion did not improve survival in patients with coronary
artery disease undergoing bypass surgery who were believed to be at high risk of sudden death on the basis of an abnormal
signal-averaged ECG and severe LV dysfunction (LVEF less than or equal to 35%).

DINAMIT (Defibrillator in Acute Myocardial Infarction Trial) (331), 674 patients with a recent MI (within 6 to 40 days), reduced LV function
(LVEF less than or equal to 35%), and impaired cardiac autonomic function (depressed heart rate variability or elevated average heart
rate) were randomized to either ICD therapy or no ICD therapy. Although arrhythmic death was reduced in the ICD group, there was no
difference in total mortality (18.7% versus 17.0%; HR for death in the ICD group 1.08; p<0.66).

AMIOVIRT (Amiodarone Versus Implantable Defibrillator in Patients with Nonischemic Cardiomyopathy and Asymptomatic Nonsustained
Ventricular Tachycardia) study (378). The trial randomized 103 patients with DCM, LVEF less than or equal to 35%, and nonsustained VT
to amiodarone or ICD. The study was stopped prematurely due to statistical futility in reaching the primary end point of reduced total
mortality (378).

DEFINITE (Defibrillators in Nonischemic Cardiomyopathy Treatment Evaluation) trial randomized 458 patients with nonischemic
cardiomyopathy, NYHA
Class I to III heart failure, LVEF less than or equal to 35%, and more than 10 premature ventricular complexes per
hour or nonsustained VT to optimal medical therapy with or without an ICD (369). With a primary end point of all-cause mortality, statistical
significance was not reached, but there was a strong trend toward reduction of mortality with ICD therapy (p<0.08). After 2 years, mortality
was 14.1% in the standard therapy group versus 7.9% among those receiving an ICD, which resulted in a 6.2% absolute reduction and a
35% relative risk reduction with ICD implantation (369). The results were consistent and comparable to those of other similar trials
(16,333,379). With the exception of DEFINITE (25% in the ICD arm), trials assessing ICD therapy in primary prophylaxis of DCM have not
generally included asymptomatic patients in NYHA functional Class I; therefore, the efficacy of ICDs in this population is not fully known.
Because mortality may be low in this subgroup, the benefit of ICD therapy is moderate at best (369).

COMPANION trial randomized patients with Class III or IV heart failure, ischemic or nonischemic DCM, and QRS duration greater than 120
milliseconds in a 1:2:2 ratio to receive optimal pharmacological therapy alone or in combination with CRT with either a pacemaker or a
pacemaker defibrillator (224). Of the 1,520 patients randomized in the trial, 903 were allocated to either the medical therapy or defibrillator
arms; of this subset, 397 (44%) had DCM. Cardiac resynchronization with an ICD significantly reduced all-cause mortality compared with
pharmacological therapy alone in patients with DCM (HR for all-cause death 0.50, 95% CI 0.29 to 0.88; p<0.015) (224).

Secondary prevention for SCD survivors, documented sustained VT/VF, syncope with inducible VT. Recurrence 20-40% in 1 year        
AVID (US), CASH (europe), CIDS (canada)

A large prospective, randomized secondary prevention trial comparing ICD therapy with Class III antiarrhythmic drug therapy
(predominantly empirical amiodarone) demonstrated improved survival with ICD therapy (319). Unadjusted survival estimates for the ICD
group and the antiarrhythmic drug group, 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
(p<0.02). Estimated relative risk reduction with ICD therapy was 39% (95% CI 19% to 59%) at 1
year, 27%
(95% CI 6% to 48%) at 2 years, and 31% (95% CI 10% to 52%) at 3 years.

(Antiarrhythmics Versus Implantable Defibrillators) (319)
CASH (Cardiac Arrest Study Hamburg) (321)
CIDS (Canadian Implantable Defibrillator Study) (322)
ICD was associated with a 50% relative risk reduction for arrhythmic death and a 25% relative risk reduction for all-cause mortality (324).

Between 73% and 83% of patients enrolled in the AVID, CASH, and CIDS trials had underlying coronary artery disease (319,321,322). The
mean LVEF ranged from 32% to 45% in these trials, which indicates prior MI in the majority of patients

Patients experiencing
cardiac arrest due to VF that occurs more than 48 hours after an MI may be at risk for recurrent cardiac arrest. If
coronary revascularization is not possible and there is evidence of significant LV dysfunction, the primary therapy for patients resuscitated
from VF should be the ICD.

sustained VT or VF is accompanied by modest elevations of cardiac enzymes, it should not be assumed that a new MI was
the cause of the sustained VT (16). Without other clinical data to support the occurrence of a new MI, it is reasonable to consider that such
patients are at risk for recurrent sustained VT or VF.
Class I for ICM: < 30% NYHA I - < 35% NYHA Class II or III

Primary electrical disorder, with most patients having 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 may be helpful (384–386). ICD implantation is recommended for selected patients with
recurrent syncope despite
drug therapy,
sustained ventricular arrhythmias, or sudden cardiac arrest (349,351,352,387,388). Furthermore, use of the ICD
for primary prevention of SCD may be considered when there is a
strong family history of SCD or when compliance or intolerance to
drugs is a concern (349,351,352,387,388). Data from genetic analysis becoming increasingly useful for clinical decision making (389–


Among selected high risk patients, the annual mortality from HCM has been estimated to be as high as 6% in reports from tertiary
centers (245,395–398). However, community-based studies suggest a more benign disease in the majority of individuals, with an
jannual mortality rate in the range of 1% or less (16,401–403). According to the 2003 ACC consensus document, the
major risk
factors include prior cardiac arrest, spontaneous sustained VT, spontaneous nonsustained VT, family history of SCD,
syncope, LV thickness greater than or equal to 30 mm, and an abnormal blood pressure response to exercise
(395). This
consensus document also noted
possible risk factors, which included AF, myocardial ischemia, LV outflow obstruction, high-
risk mutations, and intense (competitive) physical exertion
(395). The severity of other symptoms, such as dyspnea, chest pain,
and effort intolerance, has not been correlated with increased risk of SCD (16,395). An flat or hypotensive response to upright or
supine exercise testing in patients younger than 40 years old has been shown to be a risk factor for SCD, although the positive
predictive value of this finding is low (395). A normal blood pressure response identifies a low-risk group (16,395). The presence of
nonsustained VT on Holter monitoring has been associated with a higher risk of SCD, although the positive predictive accuracy is
relatively low (395). Recent analyses indicate that in a high-risk HCM cohort, ICD interventions were frequent and were highly effective
in restoring normal sinus rhythm (245). However, an important proportion of ICD discharges occur in primary prevention patients who
undergo implantation of the ICD for a single risk factor. Therefore, a single risk marker of high risk for sudden cardiac arrest may be  
sufficient to justify consideration for prophylactic ICD implantation in selected patients (245).

Patients with multiple risk factors (especially severe septal hypertrophy, greater than or equal to 30 mm) and those with
SCD (especially multiple SCDs) in close relatives appear to be at sufficiently high risk to merit consideration of ICD
therapy (16,245).
ACC ESC 2003 consensus on HCM

Risk factors that have clinical utility in identifying patients with ARVD/C who are at risk for life threatening ventricular arrhythmias
include induction of VT during electrophysiological testing, detection of nonsustained VT on noninvasive monitoring,
male gender, severe RV dilation, and extensive RV involvement
(16,341,342,345– 348,409,410). Young age at
presentation (less than 5 years), LV involvement, prior cardiac arrest, and unexplained syncope serve as markers of risk
(341,342,346 –348,411,412). Patients with genotypes of ARVD/C associated with a high risk for SCD should be considered for ICD
therapy (345).

Some authorities have proposed that an ICD should be implanted in patients with ARVD/C and an increased risk for SCD based on the
presence of a
previous cardiac arrest, syncope due to VT, evidence of extensive RV disease, LV involvement, or
presentation with polymorphic VT and RVA aneurysm, which is associated with a genetic locus on chromosome 1q42–43

ARVC Diagnosis criteria


Abnormalities in the resting ECG, including BBB or ST-segment depression, are found in most patients, but the findings do not have a
high degree of sensitivity or specificity (410,413–421). There are currently no techniques clinically useful for risk stratification for life-
threatening ventricular arrhythmias with noncompaction. Although there is no impairment of systolic function, ventricular arrhythmias
are frequent in noncompaction. Approximately 40% of children with noncompaction demonstrate complex ventricular arrhythmias.
Available clinical data indicate that sudden death is the most common cause of mortality. Although there are no prospective trials or
registry data, there are sufficient observational data to indicate that placement of an ICD as a strategy to reduce the risk of sudden
death is a reasonable clinical strategy (410,413–421).

Recommendations for ICD vary from indications to all, to only those presenting with cardiac arrest and/or EF < 30%


Brugada Page

Cardiac events such as syncope or cardiac arrest occur predominantly in the third and fourth decades of life, although presentation
with cardiac arrest in neonates or children has been reported (16,422,424). Fever can acutely predispose to cardiac arrest in the
Brugada syndrome (16,422–424.). As with long-QT syndrome, there are no data showing that family history predicts cardiac events
among family members with the Brugada syndrome (16). Accordingly, asymptomatic individuals with the characteristic ECG but with no
family history are not necessarily at low risk (16). Additionally, family members of an individual with SCD due to Brugada syndrome
should not be assumed to be at increased risk of SCD (16). Patients with a spontaneous Brugada pattern have a worse prognosis
than individuals in whom the typical ECG is observed only after pharmacological drug challenge (16,422–424).
Patients with
syncope and the ECG pattern of spontaneous ST-segment elevation have a 6-fold higher risk of cardiac arrest
patients without syncope and the spontaneous ECG pattern (16,422,424).

EPS had a low positive predictive value (23%), but over a 3-year follow-up, it had a very high negative predictive value (93%)
(16,422,424). By contrast, Priori et al. reported that electrophysiological testing has a low accuracy in predicting individuals who will
experience cardiac arrest (16,410). Priori et al. have proposed that noninvasive risk stratification based on the ECG and symptoms
provides an accurate alternative for risk stratification (16,410). Because only a single gene has been linked to the Brugada syndrome,
there is still insufficient information about the contribution of genetic defects in predicting clinical outcome (16,410,426). Specific
mutations in the SCN5A gene do not identify a subset of patients at higher risk of cardiac events (16,410,426). SCD is caused by rapid
polymorphic VT or VF that frequently occurs at rest or during sleep (16). Patients with Brugada syndrome usually do not have
ventricular extrasystoles or nonsustained runs of VT at Holter recording.


1. Survivors of cardiac arrest due to VF or hemodynamically unstable sustained VT after excluding reversible causes. (LOE: A)
2. Patient with structural heart disease and
sustained VT whether hemodynamically stable or unstable (LOE: B)
Syncope of unknown etiology with + EPS (LOE: B)
EF < 35% 40 days post MI, NYHA II-III (LOE: A)
EF < 30% 40 days post M, NYHA I (LOE: A)
EF < 40% post MI, NSVT, and + EPS (LOE: B)

1. Unexplained syncope, significant LV dysfunction, NICM (LOE: C)
Sustained VT with normal LV function (LOE: C)
HCM with 1 or more risk factors for SCD (LOE: C)
 Major: prior cardiac arrest, spontaneous sustained VT, spontaneous nonsustained VT, family history of SCD, syncope, LV thickness greater than or equal to 30 mm, and an
abnormal blood pressure response to exercise.
   Minor:  AF, myocardial ischemia, LV outflow obstruction, high-risk mutations, and intense (competitive) physical exertion

4. ARVD/C with 1 or more risk factors for SCD (LOE: C)
+ EPS, NSVT on noninvasive monitoring, male, severe RV dilation, and extensive RV involvement. Young age at presentation (less than 5 years), LV involvement, prior cardiac
arrest, and unexplained syncope. Genotypes of ARVD/C associated with a high risk for SCD should be considered for ICD therapy (locus on chromosome 1q42–43). Previous cardiac
arrest, syncope due to VT, Polymorphic VT and RVA aneurysm.

5. LQT + syncope and/or VT while on BB (LOE: B)
6. Nonhospitalizaed patients awaiting
Transplantation  (LOE: C)
Brugada + Syncope (LOE: C)
Brugada + VT (LOE: C)
CPVT + syncope and/or VT while on BB (LOE: C)
10. Reasonable for patients with
Cardiac sarcoidosis, Chagas disease (LOE: C)

1. NICM EF < 35%, NYHA I (LOE: C)
LQT + risk factors for SCD (LOE: B)
Strong family history, genetic testing, non compliance with medications        
Syncope and advanced structural heart disease without a clear cause (LOE: C)
Familial CMP associated with SCD (LOE: C)

1. ICD therapy is not recommended in patients whose expectation of survival is < 1 year (LOE: C)
2. ICD therapy is not indicated for incessant VT/VF (LOE: C)
3. Significant psychiatric history that may be aggravated by device implantation and may preclude clinical FU (LOE: C)
4. Drug refractory CHF, NYHA IV who are not candidates for CRT-D (LOE: C)
Syncope of undetermined cause without structural heart disease and with a negative EPS (LOE: C)
6. When VF/VT is amenable to catheter/surgical ablation (
AF with WPW, RV/LVOT VT, idiopathic VT, Fascicular VT) (LOE: C)
7. VF/VT in the presence of a reversible disorder (electrolyte imbalance, drugs, trauma) (LOE: B)