Signal-averaged electrocardiography has been thoroughly studied as a risk stratification tool for potentially fatal arrhythmias in patients with a previous myocardial infarction (MI). As reviewed by the Agency for Health Care Policy and Research (AHCPR) in 1998, SAECG is associated with a low positive predictive value ranging from 8%–44%, depending on the population studied.  In contrast, the negative predictive value (i.e., the ability to identify those patients who will not experience ventricular arrhythmias) ranges from 88%–97%, suggesting that the negative predictive value may be used to identify patients who would not benefit from antiarrhythmic therapy. However, a key statistic underlying the negative predictive value is the underlying prevalence of the outcome. Although sudden cardiac death is the most common cause of death in the 1-year period after infarction, it is relatively uncommon (2.5%–11.3%) and declining, due to increasing use of thrombolytic therapy, aspirin, and beta-blockers.  Thus, given the relative low incidence of arrhythmias, the high negative predictive value is not surprising.

 

In 1996, the American College of Cardiology published an expert consensus document that concluded that SAECG had an established or valuable role in clinical care in the following situations:

 

However, the ultimate validation of any diagnostic test is to determine how it is used in the management of patients, and whether the management decisions result in improved health outcomes. The following discussion focuses on the clinical use of SAECG as reported in clinical trials of antiarrhythmic therapies.

 

Over the past two decades, a large number of randomized clinical trials have evaluated the effectiveness of either antiarrhythmic drugs or AICD implantation in post-MI patients. These trials have generally used a variety of risk stratification criteria to positively select patients for intervention. By selecting patients with a sufficiently high risk of arrhythmia, the benefits of treating arrhythmia will hopefully outweigh any adverse effects of the treatment. For the purposes of this discussion, the most relevant studies are those that look at patients who have not experienced a prior episode of near fatal ventricular arrhythmia or aborted sudden death. Patients with a prior history of a potentially fatal arrhythmia are already at sufficiently high risk and are considered candidates for either antiarrhythmic therapy or AICD.

 

Initially it was thought that pharmacological suppression of premature ventricular contractions (PVCs), identified on post-MI monitoring, would reduce the incidence of subsequent sustained, symptomatic arrhythmias. The Cardiac Arrhythmia Suppression Trial (CAST) was a placebo-controlled, randomized trial that tested the efficacy of either encainide, flecainide, or moricizine in reducing arrhythmic death in patients with a lowered ejection fraction and 6 or more PVCs per hour.  CAST was terminated prematurely when an interim analysis suggested that the drug therapy was associated with an increase in the incidence in arrhythmic death. This trial raised concerns about proarrhythmic effects of antiarrhythmic drugs and has lead to caution in the use of any antiarrhythmic drug therapy.

 

The drugs in the CAST trial are known as class I antiarrhythmics, defined as those agents that slow conduction. After the failure of the CAST trial, research was focused on class III agents, which prolong repolarization. The most commonly researched member of this class of drugs is amiodarone. There have been a number of small randomized studies of amiodarone, but the largest are the EMIAT (European Myocardial Infarct Amiodarone Trial) and CAMIAT (Canadian Amiodarone Myocardial Infarction Arrhythmia Trial), both of which assessed the effect of amiodarone on mortality in patients with high-risk markers after MI.   In the EMIAT trial, patients with a history of MI were stratified according to their ejection fraction. In the CAMIAT trial, patients were recruited based on results of Holter monitoring. Therefore, neither of these key trials used SAECG as a patient selection criterion.

 

The results of both of these trials suggested that while amiodarone was associated with a decreased risk of arrhythmias, there was no overall reduction in all-cause mortality. Therefore, the major finding of these trials focused on the safety of amiodarone, in contrast to the class I agents studied in the CAST trial. The clinical effectiveness of amiodarone is less certain and may be associated with a reduction of morbidity and quality of life associated with symptomatic arrhythmias, although this outcome has not been specifically studied. It is possible that a normal SAECG could be considered to deselect patients who would be unlikely to benefit from amiodarone therapy. However, this outcome has not been specifically studied, particularly since the overall benefit of amiodarone therapy is still controversial.

 

With the somewhat disappointing results of these drug trials, attention has turned toward the use of AICDs, particularly as these devices have become more sophisticated. Early generations of AICDs required thoracotomy for insertion but miniaturization has permitted outpatient insertion with the use of local anesthesia. With this reduction in the risk associated with AICDs, there was an interest in exploring their use in patients without a prior history of sustained, symptomatic ventricular arrhythmias. Several randomized studies have now been completed. The MADIT trial recruited post-MI patients with left ventricular ejection fraction of less than 35%, non-sustained ventricular tachycardia identified on Holter monitoring or stress test, and inducible, procainide-resistant, sustained ventricular arrhythmia on electrophysiologic study (EPS).  These characteristics were thought to identify a very high risk group for ventricular arrhythmias, in part due to the desire to have very high event rates to increase the power of the trial. The MADIT trial reported a marked reduction in mortality in those receiving a defibrillator compared to patients treated conventionally, mostly with amiodarone. Following the publication of the results, the U.S. Food and Drug Administration (FDA) approved expanded labeling for defibrillators in patients who met the MADIT criteria, Medicare announced coverage for AICD in this patient population, and the American College of Cardiology (ACC) has published guidelines endorsing the study results. As noted above, SAECG was not used as a patient selection criterion, and thus is not included as a recommended test as part of the ACC guidelines.

 

In contrast to the other trials reviewed above, the CABG-Patch trial used SAECG as a positive patient selection criterion.  The CABG-Patch trial recruited patients scheduled for a CABG who had an ejection fraction of less than 36% and abnormalities on the SAECG. The use of an SAECG was based on a pilot study that showed an abnormal SAECG was associated with a mortality rate that was double that seen in those with a normal SAECG in the 2 years after CABG.  Patients were randomized to a defibrillator group or a control group, and all received CABG. After an average follow-up of 32 months, there was no evidence of improved survival among those in the defibrillator group. However, it cannot be determined whether the failure of this trial was due to the selection criteria or the treatments being compared.

 

Therefore, based on the above review, it can be seen that the SAECG has not been successfully used as a patient selection criterion in the critical randomized trials investigating both drug and device antiarrhythmic therapy in the post-MI patient. In the majority of trials, it has not been included as a patient selection criterion, and the one trial in which it was used reported negative results.

 

 

A search of the literature based on the MEDLINE database was performed for the period of 1999 through January 2006. More recent trials investigating the use of AICD in post-MI patients have not provided clarity regarding the issue of risk stratification. The MADIT-II trial selected patients solely on the basis of left ventricle ejection fraction and showed a survival benefit among those randomized to AICD.  Grimm and colleagues reported on the results of the Marburg Cardiomyopathy study, a prospective observational study designed to determine the clinical value of potential noninvasive arrhythmia risk predictors among 343 patients with idiopathic dilated cardiomyopathy and followed up for 52 +/- 21 months for major arrhythmic events.  Reduced LV ejection fraction and lack of beta blocker use were important risk factors, but results of SAECG and T-wave alternans were not. Results of SAECG were found to only be a weak predictor of sudden cardiac death in a consecutive series of 700 patients with a history of acute myocardial infarction (AMI).  In another study of 1,800 consecutive survivors of AMI who underwent reperfusion therapy, late potentials identified by SAECG were not significantly associated with the endpoints of cardiac death or serious arrhythmias. No additional data have directly linked risk stratification information provided by SAECG to improved patient outcomes, improved efficiency, or reduced costs.

 

There are inadequate data to evaluate the impact on patient management of other applications of SAECG including, but not limited to, its use in patients with cardiomyopathy; assessment of success after surgery for arrhythmia; detection of acute rejection of heart transplants; assessment of efficacy of antiarrhythmic drug therapy; assessment of success of pharmacological, mechanical, or surgical interventions to restore coronary artery blood flow; or risk stratification of patients with Brugada syndrome. Regarding the use of SAECG to identify patients with syncope who may have inducible ventricular tachycardia, even though the American College of Cardiology consensus document concluded that SAECG had an established role, data within that report reported only modest sensitivity (73%) and poor positive predictive values.  Thus the test, if used as a screening test to determine who should have electrophysiologic studies, will fail to detect many patients who have positive electrophysiologic studies.

 

A literature search was conducted using MEDLINE through June 2007. None of the articles identified would lead to a change in the policy statement. No trials were found where prospective use of SAECG was shown to improve outcomes. Thus, the policy statement remains unchanged.

 

From the 2008 Scientific Statement on risk stratification (Goldberger, 2008):

Abundant data show that an abnormal SAECG may identify patients with prior MI at risk for SCD. Given the high negative predictive value of this test, it may be useful for the identification of patients at low risk. Routine use of the SAECG to identify patients at high risk for SCD is not adequately supported at this time. Further studies are required to assess the utility of this test.

 

 

 

 

 

 

 

 

 

A literature search conducted using the MEDLINE database through August 2015 did not reveal any new information that would prompt a change in the coverage statement.