The treatment of chronic heart failure is discussed here. Acute heart failure and pulmonary edema are discussed in the next section.

A. Correction of Reversible Causes
The major reversible causes of chronic heart failure include valvular lesions, myocardial ischemia, uncontrolled hypertension, arrhythmias (especially persistent tachycardias), alcohol- or drug-induced myocardial depression, intracardiac shunts, and high-output states. Calcium channel blockers, antiarrhythmic drugs, and nonsteroidal anti-inflammatory agents are important causes of worsening heart failure. Some metabolic and infiltrative cardiomyopathies may be partially reversible, or their progression may be slowed; these include hemochromatosis, sarcoidosis, and amyloidosis. Reversible causes of diastolic dysfunction include pericardial disease and left ventricular hypertrophy due to hypertension. Once it is established that there is no reversible component, the measures outlined below are appropriate.

B. Diuretic Therapy
Diuretics are the most effective means of providing symptomatic relief to patients with moderate to severe congestive heart failure.

Few patients with symptoms or signs of fluid retention can be optimally managed without a diuretic. However, excessive diuresis can lead to electrolyte imbalance and neurohormonal activation. A combination of a diuretic and an ACE inhibitor should be the initial treatment in most symptomatic patients. When fluid retention is mild, thiazide diuretics or a similar type of agent (hydrochlorothiazide, 25-100 mg; metolazone, 2.5-5 mg; chlorthalidone, 25-50 mg; etc) may be sufficient. These agents block sodium reabsorption in the cortical diluting segment at the terminal portion of the loop of Henle and in the proximal portion of the distal convoluted tubule. The result is natriuresis and kaliuresis. These agents also have weak carbonic anhydrase inhibitor activity, which results in proximal tubule inhibition of sodium reabsorption. Thiazide or related diuretics often provide better control of hypertension than short-acting loop agents.

The thiazides are generally ineffective when the glomerular filtration rate falls below 30-40 mL/min, a not infrequent occurrence in patients with severe heart failure. Metolazone maintains its efficacy down to a glomerular filtration rate of approximately 20-30 mL/min. Adverse reactions include hypokalemia and intravascular volume depletion with resulting prerenal azotemia, skin rashes, neutropenia and thrombocytopenia, hyperglycemia, hyperuricemia, and hepatic dysfunction.

Patients with more severe heart failure should be treated with one of the loop diuretics. These include furosemide (20-320 mg daily), bumetanide (1-8 mg daily), and torsemide (20-200 mg daily). These agents have a rapid onset and a relatively short duration of action. In patients with preserved renal function, two or more doses are preferable to a single larger dose. In acute situations or when gastrointestinal absorption is in doubt, they should be given intravenously. The loop diuretics inhibit chloride reabsorption in the ascending limb of the loop of Henle, which results in natriuresis, kaliuresis, and metabolic alkalosis. They are active even in severe renal insufficiency, but larger doses (up to 500 mg of furosemide or equivalent) may be required. The major adverse reactions include intravascular volume depletion, prerenal azotemia, and hypotension. Hypokalemia, particularly with accompanying digitalis therapy, is a major problem. Less common side effects include skin rashes, gastrointestinal distress, and ototoxicity (the latter more common with ethacrynic acid and possibly less common with bumetanide).

The potassium-sparing agents spironolactone, triamterene, and amiloride are often useful in combination with the loop diuretics and thiazides. Triamterene and amiloride act on the distal tubule to reduce potassium secretion. Their diuretic potency is only mild and not adequate for most patients with heart failure, but they may minimize the hypokalemia induced by more potent agents. Side effects include hyperkalemia, gastrointestinal symptoms, and renal dysfunction. Spironolactone is a specific inhibitor of aldosterone, which is often increased in congestive heart failure and has important effects beyond potassium retention (see below). Its onset of action is slower than the other potassium-sparing agents, and its side effects include gynecomastia. Combinations of potassium supplements or ACE inhibitors and potassium-sparing drugs can produce hyperkalemia but have been used with success in patients with persistent hypokalemia.

Patients with refractory edema may respond to combinations of a loop diuretic and thiazide-like agents. Metolazone, because of its maintained activity with renal insufficiency, is the most useful agent for such a combination. Extreme caution must be observed with this approach, since massive diuresis and electrolyte imbalances often occur; 2.5 mg of metolazone should be added to the previous dosage of loop diuretic. In many cases this is necessary only once or twice a week, but dosages up to 10 mg daily have been used in some patients.

C. Inhibitors of the Renin- Angiotensin- Aldosterone System
The renin-angiotensin-aldosterone system is activated early in the course of heart failure and plays an important role in the progression of this syndrome. Inhibition of this system with ACE inhibitors should be considered part of the initial therapy of this syndrome based on their favorable effects on prognosis.

1. Angiotensin-converting enzyme inhibitors
ACE inhibitors block the renin-angiotensin-aldosterone system by inhibiting the conversion of angiotensin I to angiotensin II, producing vasodilation by limiting angiotensin II-induced vasoconstriction, and decreasing sodium retention by reducing aldosterone secretion. Because ACE is also involved in the degradation of bradykinin, ACE inhibitors result in higher bradykinin levels, which in turn stimulate the synthesis of prostaglandins and nitric oxide. Experimental data and hemodynamic studies in patients indicate that these latter actions may be important. Although the other vasodilators tend to stimulate the renin-angiotensin system and often lose part of their effect due to the resulting fluid retention, tolerance to the ACE inhibitors is uncommon.

Many ACE inhibitors are available, and at least seven have been shown to be effective for the treatment of heart failure or the related indication of postinfarction left ventricular dysfunction (see Table 11-10). ACE inhibitors reduce mortality by approximately 20% in patients with symptomatic heart failure and have been shown also to prevent hospitalizations, increase exercise tolerance, and reduce symptoms in these patients. As a result, ACE inhibitors should be part of first-line treatment of patients with symptomatic left ventricular systolic dysfunction (ejection fraction < 40%), usually in combination with a diuretic. They are also indicated for the management of patients with reduced ejection fractions without symptoms because they prevent the progression to clinical heart failure.

Because ACE inhibitors may induce significant hypotension, particularly following the initial doses, they must be started with caution. Hypotension is most prominent in patients with already low blood pressures (systolic pressure < 100 mm Hg), hypovolemia, prerenal azotemia (especially if it is diuretic induced), and hyponatremia (an indicator of activation of the renin-angiotensin system). These patients should generally be started at low dosages (captopril 6.25 mg three times daily, enalapril 2.5 mg daily, or the equivalent), but other patients may be started at twice these dosages. Within several days (for those with the markers of higher risk) or at most 2 weeks, patients should be questioned about symptoms of hypotension, and both renal function and K⁺ levels should be monitored.

ACE inhibitors should be titrated to the dosages proved effective in clinical trials (captopril 50 mg three times daily, enalapril 10 mg twice daily, lisinopril 10 mg daily, or the equivalent) over a period of 1-3 months. Most patients will tolerate these doses. Asymptomatic hypotension is not a contraindication to up-titrating or continuing ACE inhibitors. Some patients exhibit rises in serum creatinine or K⁺, but they do not require discontinuation if the levels stabilize - even at values as high as 3 mg/dL and 5.5 meq/L, respectively. Renal dysfunction is more frequent in diabetics, older patients, and those with low systolic pressures, and these groups should be monitored more closely. The most common side effects of ACE inhibitors in heart failure patients are dizziness (often not related to the level of blood pressure) and cough, though the latter is often due as much to heart failure or intercurrent pulmonary conditions as to the ACE inhibitor.

2. Angiotensin II receptor blockers
Another approach to inhibiting the renin-angiotensin-aldosterone system is the use of specific angiotensin II receptor blockers (see Table 11-10), which will block or decrease most of the effects of the system. In addition, because there are alternative pathways of angiotensin II production in many tissues, the receptor blockers may provide more complete system blockade.

However, these agents do not share the effects of ACE inhibitors on other potentially important pathways that produce increases in bradykinin, prostaglandins, and nitric oxide in the heart, blood vessels, and other tissues. The recently completed Valsartan in Heart Failure Trial (Val-HeFT) examined the efficacy of adding valsartan to ACE inhibitor therapy. No benefit in survival was observed and only a modest reduction in hospitalizations; however, in the small subset of patients without background ACE inhibitor therapy, valsartan significantly reduced both mortality and hospitalizations for heart failure. Therefore, angiotensin receptor blockers should be considered as alternatives to ACE inhibitors in ACE-intolerant patients.

3. Spironolactone
There is growing evidence that aldosterone may mediate some of the major effects of renin-angiotensin-aldosterone system activation, such as myocardial remodeling and fibrosis, as well as sodium retention and potassium loss at the distal tubules. Thus, spironolactone should be considered as a neurohormonal antagonist rather than narrowly as a potassium-sparing diuretic. The RALES trial compared spironolactone 25 mg daily with placebo in patients with advanced heart failure already receiving ACE inhibitors and diuretics and showed a 29% reduction in mortality as well as similar decreases in other clinical end points. Hyperkalemia was uncommon in this severe heart failure population, which was maintained on high doses of diuretic, but potassium levels should be monitored closely (after 1 and 4 weeks of therapy) in patients receiving ACE inhibitors. Neither the efficacy nor the safety of spironolactone has been established in the large majority of patients with mild or moderate heart failure who are taking low doses of diuretics, though this agent may be considered in patients who require potassium supplementation. A more selective aldosterone inhibitor, eplerenone, also appears effective in preventing left ventricular remodeling.

D. ß-Blockers
Although ß-blockers have traditionally been considered to be contraindicated in patients with heart failure because they may block the compensatory actions of the sympathetic nervous system, there is now strong evidence that these agents have important beneficial effects in this patient population. The mechanism of this benefit remains unclear, but it is likely that chronic elevations of catecholamines and sympathetic nervous system activity cause progressive myocardial damage, leading to worsening left ventricular function and dilation. The primary evidence for this hypothesis is that over a period of 3-6 months, ß-blockers produce consistent substantial rises in ejection fraction (averaging 10% absolute increase) and reductions in left ventricular size and mass.

Clinical trial results have been reported in nearly 14,000 patients (ranging from asymptomatic post-myocardial infarction left ventricular dysfunction to severe heart failure with left ventricular ejection fractions < 35-40%) receiving ACE inhibitors and diuretics randomized to ß-blockers or placebo. Carvedilol, a nonselective ß₁- and ß ₂-receptor blocker with additional weak a-blocking activity, was the first ß-blocker approved for heart failure in the United States after showing a reduction in death and hospitalizations in four smaller studies with a total of nearly 1100 patients. Subsequently, trials with two ß₁-selective agents, bisoprolol (CIBIS II, with 2647 patients) and sustained-release metoprolol (MERIT, with nearly 4000 patients), showed 35% reductions in mortality as well as fewer hospitalizations. Recently, a trial using carvedilol in 2200 patients with severe (New York Heart Association [NYHA] class III/IV) heart failure was terminated ahead of schedule because of a 35% reduction in mortality. In these trials, there were reductions in sudden deaths and deaths from worsening heart failure, and benefits were seen in patients with underlying coronary disease and those with primary cardiomyopathies. In all these studies, the ß-blockers were generally well tolerated, with similar numbers of withdrawals in the active and placebo groups. This has led to a strong recommendation that stable patients (defined as having no recent deterioration or evidence of volume overload) with mild, moderate, and even severe heart failure should be treated with a ß-blocker unless there is a noncardiac contraindication. In the COPERNICUS trial, carvedilol was both well tolerated and highly effective in reducing both mortality and heart failure hospitalizations in a group of patients with severe (NYHA class III or IV) symptoms, but care was taken to ensure that they were free of fluid retention at the time of initiation. In this study, one death was prevented for every 13 patients treated for 1 year - as dramatic an effect as has been seen with a pharmacologic therapy in the history of cardiovascular medicine. One trial comparing carvedilol and metoprolol (COMET) found significant reductions in all-cause mortality and cardiovascular mortality with carvedilol and thus patients may benefit from switching from metoprolol to carvedilol.

Because even apparently stable patients may deteriorate when ß-blockers are initiated, this must be done gradually and with great care. Carvedilol is initiated at a dosage of 3.125 mg twice daily and may be increased to 6.25, 12.5, and 25 mg twice daily at intervals of approximately 2 weeks. The protocols for sustained-release metoprolol use were started at 12.5 or 25 mg daily and doubled at intervals of 2 weeks to a target dose of 200 mg daily (using the Toprol XL sustained-release preparation). Bisoprolol was administered at a dosage of 1.25, 2.5, 3.75, 5, 7.5, and 10 mg daily, with increments at 1- to 4-week intervals. More gradual up-titration is often more convenient and may be better tolerated.

Patients should be instructed to monitor their weights at home as an indicator of fluid retention and to report any increase or change in symptoms immediately. Before each dose increase, the patient should be seen and examined to ensure that there has not been fluid retention or worsening of symptoms. If heart failure worsens, this can usually be managed by increasing diuretic doses and delaying further increases in ß-blocker doses, though downward adjustments or discontinuation is sometimes required. Carvedilol, because of its a-blocking activity, may cause dizziness or hypotension. This can usually be managed by reducing the doses of other vasodilators and by slowing the pace of dose increases.

E. Digitalis Glycosides
The digitalis glycosides (primarily digoxin) are the only orally active positive inotropic agents currently available. They bind to the sodium-potassium ATPase on the sarcolemmal membrane, inhibiting the sodium pump and thereby increasing intracellular sodium. This facilitates sodium-calcium exchange, with a resultant increase in cytosolic calcium, which enhances contractile protein cross-bridge formation and force generation. The digitalis glycosides also have electrophysiologic effects that may be beneficial or deleterious in individual patients. The primary therapeutic effect is an enhancement of cardiac parasympathetic tone, which delays atrioventricular conduction and reduces sinus node automaticity, thereby decreasing the ventricular response in patients with atrial fibrillation and slightly slowing the rate of patients in sinus rhythm. However, the increase in intracellular calcium and sodium may enhance automaticity of latent pacemakers, increasing the excitability of ventricular myocytes and inducing ventricular arrhythmias, especially when hypokalemia or myocardial ischemia is present.

Although the digitalis glycosides were once the mainstay of treatment of congestive heart failure, their use in patients who are in sinus rhythm has declined because they lack the benefits of the neurohormonal antagonists on prognosis and because safety concerns persist. However, their efficacy in reducing the symptoms of heart failure has been established in at least four multicenter trials that have demonstrated that digoxin withdrawal is associated with worsening symptoms and signs of heart failure, more frequent hospitalizations for decompensation, and reduced exercise tolerance. This was also seen in the 6800-patient Digitalis Investigators Group (DIG) trial, though that study found no benefit (or harm) with regard to survival. A reduction in deaths due to progressive heart failure was balanced by an increase in deaths due to ischemic and arrhythmic events. Based on these results, digoxin should be used for patients who remain symptomatic when taking diuretics and ACE inhibitors as well as for patients with heart failure who are in atrial fibrillation and require rate control.

Digoxin, the only widely employed digitalis preparation, has a half-life of 24-36 hours and is eliminated almost entirely by the kidneys. The oral maintenance dose may range from 0.125 mg three times weekly to 0.5 mg daily. It is lower in patients with renal dysfunction, in older patients, and in those with smaller lean body mass. Although a loading dose of 0.75-1.25 mg (depending primarily on lean body size) over 24-48 hours may be given if an early effect is desired, in most patients with chronic heart failure it is sufficient to begin with the expected maintenance dose (usually 0.125-0.25 mg daily). Amiodarone, quinidine, propafenone, and verapamil are among the drugs that may increase digoxin levels up to 100%. It is prudent to measure a blood level after 7-14 days (and at least 6 hours after the last dose was administered). Most of the positive inotropic effect is apparent with serum digoxin levels between 0.7 ng/mL and 1.2 ng/mL, and levels above this range may be associated with a higher risk of arrhythmias and lower survival rates, though clinically evident toxicity is rare with levels below 1.8 ng/mL. Once an appropriate maintenance dose is established, subsequent levels are usually not indicated unless there is a change in renal function or medications that affects digoxin levels or a significant deterioration in cardiac status that may be associated with reduced clearance.

Digoxin toxicity has become less frequent as there has been a better appreciation of its pharmacology, but the therapeutic-to-toxic ratio is quite narrow. Symptoms of digitalis toxicity include anorexia, nausea, headache, blurring or yellowing of vision, and disorientation. Cardiac toxicity may take the form of atrioventricular conduction or sinus node depression; junctional, atrial, or ventricular premature beats or tachycardias; or ventricular fibrillation. Potassium administration (following serum potassium measurement, since severe toxicity may be associated with hyperkalemia) is usually indicated for the tachyarrhythmias even when levels are in the normal range, but may worsen conduction disturbances. Lidocaine or phenytoin may be useful for ventricular arrhythmias, as is overdrive pacing, but quinidine, amiodarone, and propafenone should be avoided because they will increase digoxin levels. Electrical cardioversion should be avoided if possible, as it may cause intractable ventricular fibrillation or cardiac standstill. Pacing is indicated for third-degree AV block (complete heart block) and symptomatic or severe block (heart rate < 40 beats/min) if they persist after treatment with atropine. Digoxin immune fab (ovine) are available for life-threatening toxicity or large overdoses, but it should be remembered that their half-life is shorter than that of digoxin and so repeat administration may be required.

F. Vasodilators
Agents that dilate arteriolar smooth muscle and lower peripheral vascular resistance reduce left ventricular afterload. Medications that diminish venous tone and increase venous capacitance reduce the preload of both ventricles as their principal effect. Because most patients with moderate to severe heart failure have both elevated preload and reduced cardiac output, the maximum benefit of vasodilator therapy can be achieved by an agent or combination of agents with both actions. Many patients with heart failure have mitral or Tricuspid regurgitation; agents that reduce resistance to ventricular outflow tend to redirect regurgitant flow in a forward direction.

Although vasodilators that are also neurohumoral antagonists - specifically, the ACE inhibitors - improve prognosis, such a benefit is less clear with the direct-acting vasodilators. The combination of hydralazine and isosorbide dinitrate has also improved survival, but to a lesser extent than ACE inhibitors.

The intravenous vasodilating drugs and their dosages have been discussed elsewhere in this section (in the section on complications in acute myocardial infarction).

1. Nitrates
Intravenous vasodilators (sodium nitroprusside or nitroglycerin) are used primarily for acute or severely decompensated chronic heart failure, especially when accompanied by hypertension or myocardial ischemia. If neither of the latter is present, therapy is best initiated and adjusted based on hemodynamic measurements. Starting dosage for nitroglycerin is generally about 10 ug/min, which is titrated upward by 10-20 ug/min (to a maximum of 200 ug/min) until mean arterial pressure drops by 10%. Avoid hypotension (BP < 100 mm Hg systolic). For sodium nitroprusside, starting dosage is 0.3-0.5 ug/kg/min with upward titration to a maximum dose of 10 ug/kg/min.

Isosorbide dinitrate, 20-80 mg orally three times daily, has proved effective in several small studies. Nitroglycerin ointment, 12.5-50 mg (1-4 inches) every 6-8 hours, appears to be equally effective although somewhat inconvenient for long-term therapy. The nitrates are moderately effective in relieving shortness of breath, especially in patients with mild to moderate symptoms, but less successful - probably because they have little effect on cardiac output - in advanced heart failure. Nitrate therapy is generally well tolerated, but headaches and hypotension may limit the dose of all agents. The development of tolerance to chronic nitrate therapy is now generally acknowledged. This is minimized by intermittent therapy, especially if a daily 8- to 12-hour nitrate-free interval is employed, but probably develops to some extent in most patients receiving these agents. Transdermal nitroglycerin patches have no sustained effect in patients with heart failure and should not be employed for this indication.

2. Nesiritide
This new agent, a recombinant form of human brain natriuretic peptide, is a potent vasodilator that reduces ventricular filling pressures and improves cardiac output. Its hemodynamic effects resemble those of intravenous nitroglycerin with a more predictable dose-response curve and a longer duration of action. In clinical studies, nesiritide (administered as 2 ug/kg by intravenous bolus injection followed by an infusion of 0.01 ug/kg/min, which may be up-titrated if needed) produced a rapid improvement in both dyspnea and hemodynamics. The primary adverse effect is hypotension, which may be symptomatic and sustained. Because most patients with acute heart failure respond well to conventional therapy, the role of nesiritide may be primarily in patients who continue to be symptomatic after initial treatment with diuretics and nonparenteral nitrates.

3. Hydralazine
Oral hydralazine is a potent arteriolar dilator and markedly increases cardiac output in patients with congestive heart failure. However, as a single agent, it has not been shown to improve symptoms or exercise tolerance during chronic treatment. The combination of nitrates and oral hydralazine produces greater hemodynamic effects.

Hydralazine therapy is frequently limited by side effects. Approximately 30% of patients are unable to tolerate the relatively high doses required to produce hemodynamic improvement in heart failure (200-400 mg daily in divided doses). The major side effect is gastrointestinal distress, but headaches, tachycardia, and hypotension are relatively common. Angiotensin receptor blockers have largely supplanted the use of the hydralazine-isosorbide dinitrate combination in ACE-intolerant patients.

G. Positive Inotropic Agents
The digitalis derivatives are the only available oral inotropic agents in the United States. A number of other oral positive inotropic agents have been investigated for the chronic treatment of heart failure, but all have increased mortality without convincing evidence of improvement in symptoms. Intravenous agents, such as the ß1-agonist dobutamine and the phosphodiesterase inhibitor milrinone, are sometimes employed on a long-term or intermittent basis. The limited available data suggest that continuous therapy is also likely to increase mortality; intermittent inotropic therapy has never been evaluated in controlled trials, and its use is largely based on anecdotal experience. A recent randomized placebo-controlled trial of 950 patients evaluating intravenous milrinone in patients admitted for decompensated heart failure who had no definite indications for inotropic therapy showed no benefit in terms of survival, decreasing length of admission, or preventing readmission - and significantly increased rates of sustained hypotension and atrial fibrillation. Thus, the role of positive inotropic agents appears to be limited to patients with symptoms and signs of low cardiac output (primarily hypoperfusion and deteriorating renal function) and those who fail to respond to intravenous diuretics. In some cases, dobutamine or milrinone may help maintain patients who are awaiting cardiac transplantation.

H. Calcium Channel Blockers
First-generation calcium channel blockers may accelerate the progression of congestive heart failure. However, two trials with amlodipine in patients with severe heart failure showed that this agent was safe, though not superior to placebo. These agents should be avoided unless they are being utilized to treat associated angina or hypertension, and for these indications amlodipine is the drug of choice.

I. Anticoagulation
Patients with left ventricular failure and reduced ejection fractions are at somewhat increased risk of developing intracardiac thrombi and systemic arterial emboli. However, this risk appears to be primarily in patients who are in atrial fibrillation or who have large recent (within 3-6 months) myocardial infarctions. These groups should be anticoagulated. Other patients with heart failure have embolic rates of approximately two per 100 patient-years of follow-up, which approximates the rate of major bleeding, and routine anticoagulation does not appear warranted except in patients with prior embolic events or mobile left ventricular thrombi.

J. Antiarrhythmic Therapy and Implantable Cardioverter Defibrillators (ICDs)
Patients with moderate to severe heart failure have a high incidence of both symptomatic and asymptomatic arrhythmias. Although fewer than 10% of patients have syncope or presyncope resulting from ventricular tachycardia, ambulatory monitoring reveals that up to 70% of patients have asymptomatic episodes of nonsustained ventricular tachycardia. These arrhythmias indicate a poor prognosis independent of the severity of left ventricular dysfunction, but many of the deaths are probably not arrhythmia related. ß-Blockers, because of their marked favorable effect on prognosis in general and on the incidence of sudden death specifically, should be initiated in these as well as other patients with heart failure. Empiric antiarrhythmic therapy has not proved beneficial in patients with asymptomatic ventricular arrhythmias, and most other agents are contraindicated because of their proarrhythmic effects in this population and their adverse effect on cardiac function.

Patients with aborted sudden death, hemodynamically unstable ventricular arrhythmias, and unexplained cardiogenic syncope are at high risk for fatal ventricular arrhythmias. If these patients have a reasonable life expectancy and stable, nonrefractory heart failure, an implantable defibrillator is the approach of choice (again in conjunction with ß-blockade). There are not sufficient data on which to base firm recommendations for the management of patients with asymptomatic nonsustained ventricular arrhythmias. ß-Blockers are the first line of therapy, but there is no evidence that antiarrhythmic drugs are beneficial, and they carry substantial risk (with the exception of amiodarone). In the second Multicenter Automatic Defibrillator Implantation Trial (MADIT II), 1232 patients with prior myocardial infarction and an ejection fraction = 30% without a history of symptomatic ventricular arrhythmias were randomized to an ICD or a control group. Mortality was 31% lower in the ICD group, which translated into nine lives saved for each 100 patients who received a device and were followed for 3 years. The potential implications of this result are enormous, since this translates into an estimated cost of $300,000 per life saved. Before acting on this result, corroborative data should be awaited from the ongoing Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT). Patients in this larger trial appear to be more symptomatic and more aggressively treated. Importantly, it includes a third arm in which patients have been randomized to amiodarone.