A recent analysis of the ICD Registry from the National Cardiovascular Data Registry raises significant concerns about the effectiveness of the treatment guidelines for implantable cardioverter defibrillators for the primary prevention of sudden cardiac arrhythmic death.
That study indicates that, although the guidelines as proposed by the sponsoring societies were adhered to in most implantations, a disappointing one-quarter of the implantations were outside of the guidelines. Of the over 111,707 ICDs implanted between 2006 and 2009, 25,145 (22.5%) were implanted in patients who were outside of the recommended guidelines. Certainly one can make a case for the fact that these are only guidelines, and doctor should have the prerogative to make clinical decisions, but there have been few guidelines that have been as carefully explored as those for ICD implantation. They emphasize not only the benefit of the devices implanted within the guidelines, but the hazards of the implantation outside of the guidelines (JAMA 2011;305:43-9).
Four guideline deviations were identified and included implantation carried out within 40 days of an acute MI, in 37%; within 3 months of coronary artery bypass surgery, in 3%; in patients with New York Heart Association class IV symptoms, in 12%; and in newly diagnosed heart failure, in 62%. There are adequate randomized clinical trials that clearly show the lack of benefit or increased risks of implantation in these four classes of patients.
Both the cardiology community and the device manufacturers have emphasized the importance of ICD therapy in as many individuals who fit the criteria for implantation as possible. Despite this effort, the implantation of ICDs on the potential candidates with systolic heart failure has significantly lagged. These results may have a further cooling effect on the rate of future implantation.
There are adequate randomized clinical trials that clearly demonstrate the lack of benefit or increased risks of implantation in these four classes of heart failure patients. The delay in implantation in new heart failure patients has been dictated by the observations that many individuals improve cardiac function after an acute event. In addition, a number of studies have shown that implantation at time of surgery is without merit and with some risk, and we have learned that the mode of death in NYHA class IV patients is dominated by progressive heart failure and not primary arrhythmias.
The use of ICDs within 40 days of an acute MI is of particular importance. Several studies have raised concern about the increase in heart failure mortality in patients who have experienced both appropriate and inappropriate ICD discharge for arrhythmias. It is unclear whether the increased heart failure precedes or is a result of ICD discharge. Similar observations were made in patients in whom an ICD was implanted early after an acute MI for the primary prevention of arrhythmic deaths. A recent study of the Defibrillation in Acute Myocardial Infarction Trial (DINAMIT) re-examines the observation that that heart failure mortality increased in patients who received an ICD shock (Circulation 2010;122:2645-52). DINAMIT enrolled patients 6–40 days after an acute MI in patients with left ventricular ejection fraction of less than 35%. Although ICD therapy decreased arrhythmic deaths, there was an increase in heart failure mortality in those patients who had an ICD shock, which resulted in a net increased mortality.
Both the American College of Cardiology and the Heart Rhythm Society have expressed concern about the recent report and plan to further emphasize ICD guidelines. The HRS is cooperating with the Department of Justice in an investigation of the inappropriate Medicare payments of ICDs “to lend expertise concerning the proper guidelines for clinical decision making,” according to a statement. Should the Justice Department become involved in the appropriate use of medical guidelines, an entirely new disturbing dimension would be introduced in regard to guideline application. In the meantime, both cardiologists and noncardiologists should rethink their decisions in regard to the use of ICDs. They clearly represent an important medical advance that has saved the lives of many of our patients, but their use does have significant risks which must be balanced against their benefit.