Current issues of ACP Journal Club are published in Annals of Internal Medicine


Implantable cardioverter-defibrillators used at the time of bypass surgery did not improve survival among high-risk patients

ACP J Club. 1998 May-June;128:61. doi:10.7326/ACPJC-1998-128-3-061

Source Citation

Bigger JT Jr, for the Coronary Artery Bypass Graft (CABG) Patch Trial Investigators. Prophylactic use of implanted cardiac defibrillators in patients at high risk for ventricular arrhythmias after coronary-artery bypass graft surgery. N Engl J Med. 1997 Nov 27;337:1569-75.



To evaluate the effect of a prophylactic implantable cardioverter-defibrillator (ICD) inserted at the time of coronary artery bypass graft (CABG) surgery on all-cause mortality in patients at high risk for sudden death.


Randomized controlled trial with a mean follow-up of 32 months (CABG Patch Trial).


35 centers in the United States and 2 centers in Germany.


900 patients < 80 years of age (mean age 63 y, 84% men) who were scheduled for CABG, had a left ventricular ejection fraction < 0.36, and had abnormalities on a signal-averaged electrocardiogram. Exclusion criteria were history of sustained ventricular tachycardia (VT) or fibrillation, diabetes mellitus with poor control of blood glucose levels or recurrent infections, previous or concomitant aortic or mitral valve surgery, concomitant cerebrovascular surgery, serum creatinine level > 265 mmol/L, emergency CABG, or noncardiac conditions with a life expectancy of < 2 years.


After surgery, patients were stratified by center and by left ventricular ejection fraction (≤ 0.20 or 0.21 to 0.35) and allocated to ICD therapy (n = 446) or to no defibrillator therapy (n = 454).

Main outcome measures

All-cause mortality. Secondary outcomes were complications after surgery and adverse events during long-term follow-up.

Main results

Analysis was by intention to treat. The groups did not differ for total deaths (101 deaths in the ICD group vs 95 in the group without defibrillators, P = 0.64). 71 (70%) deaths among patients in the ICD group and 72 (76%) deaths in the control group were from cardiac causes. The hazard ratio comparing the risk for death per unit of time in the 2 treatment groups was 1.07 (95% CI 0.81 to 1.42). The results remained similar after adjustment for baseline variables. After surgery, more infections were reported among patients in the ICD group (P < 0.05), and during long-term follow-up, more myocardial infarctions were reported among patients in the control group (P < 0.05).


Prophylactic implantation of a cardioverter-defibrillator at the time of coronary artery bypass graft surgery did not improve survival of patients at high risk for sudden death.

Sources of funding: National Heart, Lung, and Blood Institute and Guidant/Cardiac Pacemakers, Inc.

For article reprint: Dr. J.T. Bigger, Jr, Data Coordinating Center, The CABG Patch Trial, PH103-D, Columbia University, 630 West 168th Street, New York, NY 10032, USA. FAX 212-305-7141.


Implantable cardioverter-defibrillators reduced mortality in patients resuscitated from ventricular arrhythmias

The AVID study provides the first published results of a randomized clinical trial comparing ICD with the best available anti-arrhythmic drug therapy for patients who had ventricular fibrillation or symptomatic sustained VT. All-cause mortality was 39% less at 1 year, and the benefit persisted at 2 and 3 years. Minor reservations about the validity of the results may be raised because the ICD group had a lower prevalence of congestive heart failure at baseline and used more β-blockers, but adjusted analyses did not significantly change the outcomes. No standard loading regimen was set for amiodarone, and the dose at 2 and 3 years is rather conservative for such a high-risk population. Nevertheless, the evidence for reduced mortality with the ICD was strong. Serious side effects were relatively rare with both ICD (occurring mainly in association with implantation) and amiodarone. Very few patients were randomly allocated to sotalol.

Among patients who have had life-threatening arrhythmias, the ICD seems to provide longer survival than empiric amiodarone. Is ICD therapy affordable? Cost-effectiveness studies are rare, but early studies indicate costs in the range of U.S. $115 000/life-year saved (1)—a range considered to be problematic even for a wealthy society. However, only relatively short follow-up is available in the AVID study and the survival curves do not seem to be converging. With improving technology, longer battery life, and decreasing costs, the cost-benefit analysis may eventually become more favorable.

The CABG Patch Trial compared the ICD with a no-defibrillator control in a population of patients who were having CABG, who had depressed ejection fraction, and who had an abnormal electrocardiogram, placing them at substantial risk for arrhythmic death. After a mean follow-up of 32 months, all-cause mortality rates in the 2 groups were virtually identical and were unchanged by adjustment for prespecified covariants. The use of cardiac drugs, including antiarrhythmic drugs and β-blockers, was similar in the 2 groups.

The Multicenter Automatic Defibrillator Implantation Trial (MADIT) (2) enrolled patients with nonsustained VT, low ejection fraction, and nonsuppressible VT on electrophysiologic testing. The patients had a surprisingly high mortality rate that was greatly reduced by the ICD.

For the CABG Patch Trial, it seems that the deaths of the persons were less able to be prevented by the ICD than were those of patients in the AVID trial and MADIT. Although the revascularization may have decreased the risk for ischemia-predicted arrhythmias, the rate of ICD discharge was relatively high and similar to that in MADIT, an observation that is difficult to explain if fewer potentially lethal arrhythmias were present. Data on the cause of death and from the stored electrocardiograms (available only in the latter part of the study) may help to clarify this possibility.

For now, which patients should receive an ICD? Those who have had ventricular fibrillation or sustained symptomatic VT (the AVID population) should receive an ICD, provided that arrhythmias do not recur frequently and that patients are free of severe uncontrolled ischemic heart disease or congestive heart failure and other illnesses associated with a high risk for short-term mortality. Every effort should be made to optimize medical therapy, especially the management of myocardial ischemia and congestive heart failure. Although the cost per life-year saved is high, the therapy is substantially more effective than optimal antiarrhythmic drug therapy and the cost is likely to decrease. Although the MADIT patients with nonsustained VT, nonsuppressible VT, and low ejection fraction benefited from the ICD, using this treatment is likely to be very costly. In addition, the results of a single small trial are, in my opinion, insufficient to justify this approach over medical therapies for now. Further data from studies, such as MADIT II, may provide clearer therapeutic directions. Patients with a low ejection fraction and other predictors of risk for life-threatening arrhythmias but no VT or symptomatic ventricular arrhythmia do not seem to benefit from ICD implantation.

John A. Cairns, MD
University of British ColumbiaVancouver, British Columbia, Canada


1. Larsen GC, McAulty JH, Hallstrom A, et al. Hospitalization changes in the antiarrhythmics versus implantable defibrillators (AVID) trial: the AVID economic analysis study [Abstr]. Circulation. 1997;96:1-77.

2. Moss AJ, Hall WJ, Cannon DS, et al. Improved survival with an implanted defibrillator in patients with coronary disease at high risk for ventricular arrhythmia. N Engl J Med. 1996;335:1933-40.