Dr. Weintraub argues that eplerenone, an aldosterone blocker, is more cost-effective than the ICD. Although his estimated incremental cost-effectiveness ratio of $10 402 per life-year gained for eplerenone seems much more favorable than our base-case ratio of $50 500 per life-year gained for an ICD, we caution against these comparisons when the interventions are not compared head-to-head. Both MADIT-II and EPHESUS enrolled appreciably different patients. Whereas all patients in EPHESUS had clinical or radiographic evidence of heart failure, half of the patients in MADIT-II did not. All patients in EPHESUS were enrolled 3 to 14 days after MI, but 88% of patients in MADIT-II had had an MI more than 6 months before enrollment. These differences make comparisons between eplerenone and ICD therapy difficult (1, 2).

Another reason to caution against such comparisons is that calculation of the incremental cost-effectiveness ratio depends on many assumptions. Unless these assumptions are identical, ratio comparisons are not valid. Among many other assumptions, the time horizon, the change of hazard over time, costs of both interventions, and intensity of follow-up visits are important. It is not clear to us what assumptions were used in the cost-effectiveness ratio calculated by Dr. Weintraub. On the basis of results of the Defibrillator in Acute Myocardial Infarction Trial (DINAMIT), Dr. Weintraub states that ICDs do not reduce mortality within the first month after MI. Although we agree with Dr. Weintraub's interpretation, we believe this question cannot be settled with 1 clinical trial—particularly because patients in the control group underwent coronary revascularization much more frequently than those in the ICD group during the course of DINAMIT (3). A recent analysis of the Valsartan in Acute Myocardial Infarction Trial (VALIANT) showed the risk for sudden death to be highest in the first 30 days after MI in patients with left ventricular dysfunction, heart failure, or both (4). Therefore, the question of whether an ICD is beneficial shortly after an MI deserves further examination.

Notwithstanding the ambiguities in Dr. Weintraub's letter, we agree with him on the following point: When pharmacotherapy is life-saving and cost-effective, society should be encouraged to make such therapy available to those who need it. Dr. Jafary brings up a good point regarding the limited access to ICDs in developing countries and the need for cheaper devices; however, affordability is a separate issue from cost-effectiveness. It is up to each country to decide how much of its wealth to invest in health care.