Early Initiation of β Blockade in Heart Failure: Issues and Evidence

Abstract

Despite clinical trials demonstrating that inhibitors of the renin‐angiotensin and sympathetic nervous systems can reduce the mortality and morbidity risk associated with heart failure, these drugs have remained underutilized in general clinical practice. In particular, many patients with heart failure due to left ventricular systolic dysfunction fail to receive β blockers, although this class of drugs, as well as other antihypertensive agents such as angiotensinconverting enzyme inhibitors or angiotensin receptor blockers, are recommended as part of routine heart failure therapy by national expert consensus guidelines. In‐hospital initiation of β‐blocker therapy may improve long‐term utilization by physicians and compliance by patients through obviating many of the misperceived dangers associated with β blockade. The following review of the clinical trial data from the Randomized Evaluation of Strategies for Left Ventricular Dysfunction (RESOLVD) trial, the Metoprolol Controlled‐Release Randomized Intervention Trial in Heart Failure (MERIT‐HF), the Cardiac Insufficiency Bisoprolol Study II (CIBIS‐II), the Carvedilol Prospective Randomized Cumulative Survival (COPERNICUS) trial, and the Initiation Management Predischarge Process for Assessment of Carvedilol Therapy for Heart Failure (IMPACT‐HF) trial on the efficacy, safety, and tolerability of β blockers indicates that early initiation can be safely achieved and can improve patient outcomes.

Numerous clinical trials demonstrate that inhibitors of the renin‐angiotensin‐aldosterone and sympathetic nervous systems can reduce the occurrence of and the mortality and morbidity risk associated with heart failure (HF). This type of pharmacologic therapy has, nonetheless, remained underutilized in general clinical practice. ¹ , ² Despite the effectiveness of these agents in lowering blood pressure (BP) and reducing the occurrence of HF, hypertension remains the second most common cause of HF. The use of angiotensin‐converting enzyme (ACE) inhibitors in HF is progressively increasing, but β blockers remain underprescribed, even though they are recommended as part of routine HF therapy by national expert consensus guidelines. ³ , ⁴ A recent university hospital study found that while HF was treated with an ACE inhibitor or an angiotensin receptor blocker (ARB) in 73% of cases, β blockers were administered to only 28% of patients. Of particular concern, noncardiologists were found to prescribe a β blocker for HF only half as often as cardiologists, or in 16% of cases. ⁵ This underutilization persists despite the publication of large, placebo‐controlled, randomized clinical trials demonstrating a significantly greater reduction in mortality and morbidity risk in HF patients when β blockers were added to ACE inhibitor therapy. ⁶ , ⁷ , ⁸ A recent European study found that while 60% of physicians reported awareness of these benefits, the same physicians prescribed these agents to only about one third of their HF patients. ⁹ In addition to withholding β blockers in HF, their use is often delayed. Withholding or delaying β blockade in HF is based on several clinical misapprehensions.

First, there is a general fear that decreasing adrenergic support from an already failing heart may further worsen cardiac function, potentially leading to pulmonary edema and/or cardiogenic shock. ¹⁰ , ¹¹ This concern has led to the recommendation that β‐blocker therapy not be initiated until 2‐4 weeks after an HF hospitalization discharge to allow time for patients to stabilize on standard medications. ⁴ Such recommendations contribute to the underutilization of β‐blocker therapy. There are higher rates of β‐blocker use 1 year following hospitalization in patients who initiated therapy before discharge (57%), compared with those who did not (18%). ¹²

A second major reason for delaying β blockade is the belief that the benefits of β blockade do not become apparent for several months, based on the finding that quantitative measurements of left ventricular (LV) function do not improve before then. ³ , ¹¹ This has promoted a sense of complacency and lack of urgency to begin therapy when patients with LV systolic dysfunction (LVSD) appear stable and free from clinical evidence of HF. As concluded by Gattis et al., ¹³ since clinicians are reluctant to begin a new therapy in patients who are stable and are reticent to initiate β blockade in patients with clinical HF, “the likelihood that treatment will ever be started is fairly low.” A recent registry of patients admitted for HF in select hospitals in the United States reported that only 48% of patients presenting with prior HF had been receiving a β blocker, compared with 70% who had been prescribed diuretics. ¹⁴ Implementing β blockade in HF is necessarily a slow process requiring stepwise titration, careful patient evaluation, and the possible readjustment of other cardiovascular (CV) medications. This may be considered too difficult, complicated, cumbersome, or time consuming for the busy internist or general practitioner. ¹⁵

This paper will review clinical trial data demonstrating that early use of β blockers can be safely achieved in a wide range of patients with symptomatic HF, and that mortality and morbidity benefits begin to be observed as early as 2–3 weeks after initiation of therapy. These findings should encourage physicians to increase the use of β blockers in most HF patients, with the expectation of reducing death and retarding HF progression.

HYPERTENSION, HF, AND β BLOCKERS

Hypertension is an important risk factor for CV disease morbidity and mortality. Risks of coronary heart disease, stroke, HF, and CV disease are positively correlated with increases in systolic BP (SBP) in both men and women. ¹⁶ , ¹⁷ Hypertension is often observed with other coronary heart disease risk factors, such as dyslipidemia, diabetes mellitus, renal disease, and obesity, thus increasing the risk of CV events. ¹⁸

A review of major long‐term hypertension treatment trials in over 13,000 subjects found that active antihypertensive treatment was associated with 54% fewer HF cases than placebo treatment. ¹⁹ In the 4.5‐year Systolic Hypertension in the Elderly Program (SHEP) ²⁰ (SBP >160 mm Hg, diastolic BP <90 mm Hg), there was a 49% relative risk reduction for developing fatal or nonfatal HF in treated patients compared with placebo patients (p<0.001). Patients with evidence of a prior myocardial infarction (MI) had an 89% lower risk of developing HF with active antihypertensive treatment compared with placebo (p=0.002).

Among persons who develop HF, 70% of men and 77% of women have preexisting hypertension. ²¹ Because of the direct relationship between hypertension and LV remodeling, ²² the American Heart Association/American College of Cardiology considers all persons with hypertension to be at risk of HF. ²³ Diuretics, ACE inhibitors, ARBs, β blockers, and calcium channel blockers are recommended for the treatment of comorbid conditions that are commonly present in patients with hypertension. ¹⁹ , ²⁴ , ²⁵ , ²⁶ , ²⁷ , ²⁸ , ³⁰ , ³¹ , ³² , ³³ , ³⁴

DATA ON INITIATION PERIOD: EFFICACY, SAFETY, AND TOLERABILITY OF β BLOCKERS

Several large, randomized, placebo‐controlled clinical trials have evaluated outcomes for the β blockers metoprolol, bisoprolol, and carvedilol when added to standard treatment comprised of ACE inhibitors and diuretics, with or without digoxin, in patients with mild‐to‐severe HF. These serve as evidence for recommending routine β‐blocker use in HF ⁶ , ⁷ , ⁸ These data support the efficacy, safety, and tolerability of beginning β blockers early in patients presenting with clinical signs and symptoms of HF due to LVSD.

RESOLVD

The Randomized Evaluation of Strategies for Left Ventricular Dysfunction (RESOLVD) ³⁵ trial assessed the effects of adding metoprolol succinate or placebo to symptomatic HF patients who had received treatment with either an ACE inhibitor (enalapril), an ARB (candesartan), or both for 5 months (plus a diuretic in 84% of patients). Patients were symptomatic, with a mean ejection fraction (EF) <0.29, and 10%–16% had an HF etiology other than ischemic heart disease or idiopathic cardiomyopathy. After 24 weeks of added treatment, patients randomized to metoprolol experienced approximately three times as many hospitalizations due to HF, but had a corresponding 53% decrease in mortality compared with a regimen that did not include a β blocker. Importantly, adding metoprolol to either enalapril, candesartan, or both resulted in significantly improved LV volumes and EF. There were no significant mean BP differences among the treatment groups at baseline (122/74 mm Hg) or at 43 weeks; BP decreased similarly in all treatment groups by a mean of −3 to −10 mm Hg SBP and −2 to −5 mm Hg diastolic BP. ³⁶

MERIT‐HF

The Metoprolol Controlled‐Release Randomized Intervention Trial in Heart Failure (MERIT‐HF) ⁶ was a randomized, placebo‐controlled trial of metoprolol succinate added to standard therapy in nearly 4000 New York Heart Association class II–IV HF patients with an EF <0.40 who were followed for a mean of 1 year. Hypertension was prevalent in 44% of patients and 35% of patients had a nonischemic HF etiology. Although there were no mean baseline BP differences (130/78 mm Hg), metoprolol did decrease mean SBP compared with placebo (−2.1 vs. 3.5 mm Hg; p=0.013). Overall, there was little worsening of HF during early treatment. The risk of discontinuation specifically due to worsening HF was 25% lower in metoprolol patients compared with placebo. ⁶ , ³⁷

While most patients reported no change in dyspnea or fatigue during the titration period, worsening HF symptoms were reported in 5.2% of patients randomized to metoprolol compared with 4.6% receiving placebo. ³⁷ However, more patients with severe LVSD (EF <0.25) who received metoprolol reported greater improvement and less deterioration in symptoms than in the non‐β‐blocker group. ³⁷ Bradycardia was the most common reason for not reaching metoprolol target doses within the first 90 days. ³⁷ Early mortality rates were similar in the two randomized groups. During titration, the Kaplan‐Meier curves for the combined end point of all‐cause mortality or all‐cause hospitalization were similar for all patients studied; however, after the first 60 days, the curves diverged, with fewer events occurring in the metoprolol group. ³⁷ Therapy was well tolerated, as shown by the overall low rate of early discontinuation. Fewer patients receiving metoprolol died or required hospitalization for worsening HF than those receiving placebo.

CIBIS‐II

Bisoprolol or placebo was randomly assigned to 2647 New York Heart Association class III—IV patients with an EF ≤0.35 despite receiving standard therapy with an ACE inhibitor and diuretic, with or without digoxin, in the Cardiac Insufficiency Bisoprolol Study II (CIBIS‐II) ⁷ trial. About 39% of patients had an undefined HF etiology and mean baseline BP was 130/80 mm Hg. The CIBIS‐II patient population appeared to have somewhat more advanced HF than the patients in MERIT‐HF, with an annual mortality of 13.2% compared with 11.0% in the placebo groups. Nevertheless, the relative risk of death associated with treatment using bisoprolol or metoprolol compared with placebo was equivalent (0.66) and the relative risk of sudden death was also comparable: 0.59 and 0.56 for metoprolol and bisoprolol, respectively. ⁷ , ³¹ There were significantly fewer hospital admissions for worsening HF with β blockers in both studies, and both trials were terminated early due to significant mortality advantages in treated patients. ⁶ , ³¹ Unlike MERIT‐HF, there were no differences in treatment discontinuations between β blockers and placebo.

COPERNICUS

The Carvedilol Prospective Randomized Cumulative Survival (COPERNICUS) ³⁸ trial recruited patients with severe chronic HF, i.e., symptoms at rest or on minimal exertion and an EF <0.25. A nonischemic HF etiology was present in 33% of patients. Mean baseline BP was 123/76 mm Hg and there was no significant SBP difference between treatments at the trial's conclusion. This population suffered from more advanced HF than either MERIT‐HF or CIBIS‐II, as indicated by the higher annual mortality rate of 19.7% in the placebo group. Patients were also required to be under treatment with a diuretic and either an ACE inhibitor or an ARB. COPERNICUS assessed the efficacy and safety of carvedilol, a nonselective β blocker with α₁‐blocking action.

Patients (N=2259) were randomly allocated to begin either carvedilol or placebo treatment as soon as they were clinically euvolemic, and the study medication was titrated from 3.125 mg to 25 mg b.i.d. over a minimum of 6 weeks as tolerated; two thirds of patients were able to achieve full target doses of carvedilol by 12 weeks. During this period, 21% fewer patients discontinued carvedilol than placebo.

Overall, serious adverse events occurred less frequently with carvedilol than with placebo. Worsening HF was reported in fewer, rather than more, patients taking carvedilol (17% vs. 24%). MI, unstable angina, and ventricular fibrillation occurred significantly less often with carvedilol. The risk of the combined end point of death or hospitalization was reduced by 24% in the carvedilol group. These results led to the premature termination of the COPERNICUS study. ³⁹ , ⁴⁰

Unlike the findings in MERIT‐HF, benefits appeared as early as 14–21 days following initiation of treatment and continued throughout the length of the study. In addition, carvedilol was associated with similar benefits in the highest‐risk patient subgroup (EF <0.15). Kaplan‐Meier curves for all‐cause mortality and for the combined end point of death, hospitalization, or discontinuation began to separate within the first 2 weeks for all randomized patients as well as for the high‐risk subgroup and continued to diverge throughout the entire 8 weeks of early treatment (Figure 1). ⁴¹ Early β blockade with this nonselective β blocker with α₁‐blocking action, added to conventional therapy—even in the setting of severe or recent decompensation—resulted in additional clinical benefit without the risk of worsening HE.Furthermore, these benefits appeared within 2 weeks—without an initial period of worsening—and are similar to those seen for long‐term β‐blocker therapy in HE It therefore does not seem prudent to wait 2–4 weeks after hospitalization to initiate carvedilol, as recommended for β blockers in practice guidelines.

Figure 1.

Kaplan‐Meier risk analysis for all randomly assigned and high‐risk patients (ejection fraction ≤0.15) from COPERNICUS. Acronym is expanded in the text. Adapted with permission from JAMA. 2003;289:712‐718.⁴¹

IMPACT‐HF

Further evidence challenging guideline recommendations to delay β‐blocker initiation after hospital discharge comes from the Initiation Management Predischarge Process for Assessment of Carvedilol Therapy for Heart Failure (IMPACT‐HF) ⁴² trial. The study tested the hypothesis that patients who were started on β‐blocker therapy during hospitalization would be more likely to be maintained on that therapy as outpatients. The trial compared the end point of β‐blocker use 60 days following hospital discharge in 363 patients with LVSD who had been started on carvedilol predischarge compared with any β blocker started 2 weeks or more after discharge according to usual care. Of the carvedilol‐initiated patients, 29% had a nonischemic HF etiology and 61% had hypertension, compared with 22% and 64%, respectively, of patients initiated on any β blocker. Mean BP was similar in carvedilol‐treated and any β‐blocker‐treated patients at baseline (136/80 mm Hg) and at discharge (120/70 mm Hg).

IMPACT‐HF found that 91% of patients who were started on carvedilol in the hospital were still receiving it 60 days after discharge. In contrast, 18% fewer patients (73%; p<0.0001) who did not have β‐blocker therapy initiated in the hospital were receiving β blockers. ⁴² The combined end point of death or rehospitalization began to diverge within 2 weeks of randomization, with lower mortality in the predischarge carvedilol initiation group. This early benefit was not accompanied by an increase in other serious adverse events or an increase in hospital length of stay. IMPACT‐HF showed that beginning β‐blocker therapy in the hospital improves long‐term maintenance, and COPERNICUS demonstrated the safety—as well as the mortality and morbidity outcome benefits—of early initiation of β‐blocker therapy.

Choice of β Blocker

Early studies have demonstrated that the nonselective β blockers propranolol,timolol,and carvedilol reduced death and reinfarction when administered to patients following acute MI (AMI). ³³ , ³⁴ , ⁴⁴ In the β‐Blocker Heart Attack Trial (BHAT), propranolol or placebo was given randomly to AMI patients 5–21 days after admission. During the first 30 days of treatment, propranolol‐treated patients with prior HF experienced a greater incidence of recurring HF than those receiving placebo; most of this difference occurred early (during the first 2 weeks of study medication) but did not persist. ⁴³ Interestingly, BHAT found that although propranolol decreased total mortality regardless of HF history, it decreased sudden death significantly more in patients with, than without, prior HF (47% vs. 13%, respectively). ⁴⁵ The Carvedilol Post‐Infarct Survival Control in LV Dysfunction (CAPRICORN) ³³ trial found that post‐MI patients with LV dysfunction benefited significantly with the use of carvedilol given at 3–21 days after the MI. All‐cause mortality was significantly reduced by 23% with no more HF adverse events than placebo.

Use of β₁‐selective blockers, such as metoprolol, has become part of recommended therapy in AMI in recognition of their significant effect on mortality reduction with less risk of worsening HF. Studies with the early administration of IV metoprolol have shown a one third reduction in overall mortality risk compared with placebo. ⁴⁶ Unlike propranolol, the marked clinical benefits after AMI extended to patients who had presented with signs of HF at admission; these patients, in fact, benefited most from early use of β₁‐selective blockers. ⁴⁷

Differences in pharmacology may influence the early initiation of β‐blocker therapy for HF, as it does for AMI. While β₁‐selective blockers have been proven to reduce mortality, they may temporarily cause initial decreases in cardiac output and renal blood flow and increases in vascular resistance. ⁴⁸ Carvedilol, a nonselective β blocker, has vasodilatory activity which maintains cardiac output and may decrease vascular resistance. This activity may protect the patient from worsening hemodynamic changes in the early stages of initiation. ⁴⁸ In addition, by blocking β₁β₂, and α₁ adrenergic receptors, the effects of catecholamines may be reduced to a greater degree than with β₁‐specific blockers. ⁴⁹

The clinical correlate of the pharmacologic differences between carvedilol and metoprolol has been demonstrated in a recent head‐to‐head comparison in HF patients. The Carvedilol or Metoprolol European Trial (COMET) ⁵⁰ randomized 3029 patients with chronic New York Heart Association class II–IV HF and EF <0.35 to either carvedilol or metoprolol tartrate for up to 3.5 years. Hypertensive HF was present in 18% of patients; 37% of patients had hypertension, and mean baseline BP was 126/77 mm Hg. Carvedilol was started at 3.125 mg b.i.d. and titrated to 25 mg b.i.d. Metoprolol tartrate was started at 5 mg and target dose was 50 mg b.i.d. After a mean follow‐up period of 58 months, allcause mortality was reduced by 17% (p=0.0017) and there was a 20% reduction (p=0.0004) in the risk of CV death with carvedilol compared with metoprolol. Although there were significant BP differences at 4 months and differences throughout the trial, a multivariable analysis showed that the superiority of carvedilol was not affected by SBP. ⁵¹ The risk of death or hospitalization for worsening HF was also decreased by 11% (p=0.02) with carvedilol compared with metoprolol tartrate. ⁵² The incidence of discontinuations and hospital admissions, however, was not significantly different.

Other factors may influence the choice of β blocker for use in HF patients. Because diabetes and atherosclerosis are commonly associated with HF, the metabolic effects of drugs used to treat HF should also be considered. Atherosclerotic disease is the cause of HF in at least half of all patients enrolled in HF trials, such as CIBIS‐II and MERIT‐HF. ⁶ , ⁷ HF is also associated with an increased risk of diabetes⁵³; about 25% of patients recruited for large clinical trials in HF have a history of diabetes. ⁵⁴

The effects of metoprolol tartrate and carvedilol on insulin sensitivity and lipoprotein profiles were compared in a study of 72 nondiabetic hypertensive patients. After 12 weeks of treatment, insulin sensitivity measured by the glucose metabolic clearance rate decreased 14% in patients receiving metoprolol, but increased 9% in the carvedilol group. Metoprolol therapy was also associated with a more atherogenic lipid profile marked by a decrease in high‐density lipoproteins and an increase in triglycerides; these parameters were unchanged in patients receiving carvedilol. This suggests that the possible effects of β blockade on carbohydrate and lipid metabolism are being offset by the concomitant α₁‐blocking activity of carvedilol. ⁵⁵ Likewise, in hypertensive patients with diabetes, treatment with the β₁‐selective blocker atenolol for 24 weeks resulted in a rise in fasting glucose and insulin levels—but a decrease with carvedilol. In COMET, there was a 22% decrease in new‐onset diabetes‐related adverse events with carvedilol compared with metoprolol tartrate. ⁵² The recently published Glycemic Effects in Diabetes Mellitus: Carvedilol‐Metoprolol Comparisan in Hypertension (GEMINI) trial showed a benefit on metabolic parameters with the use of carvedilol compared with metoprolol tartrate in patients with diabetes and hypertension. Glycosylated hemoglobin and insulin resistance worsened in the metoprolol tartrate‐treated patients, but was unchanged in the carvedilol‐treated patients. There were no differences in mean BP at baseline (149/87 mm Hg) or at the end of the trial (132/77 mm Hg). ⁵⁶ In addition, carvedilol also possesses potent antioxidant activity, an effect that may protect against loss of cardiac myocytes in chronic HF, especially in the large proportion of cases due to ischemic cardiomyopathy. ⁵⁷

Ventricular Function and Cardiac Remodeling in Early β Blockade

The proven chronic (more than 1 month) benefits of β blockade in HF have been ascribed to its effects on intrinsic systolic function related to enhanced myocyte contractility, and may underlie the longer‐term reversal of ventricular remodeling that results from cardiac failure, including ventricular chamber enlargement, hypertrophy, and increased sphericity. ⁴⁸

The effect of early‐phase metoprolol use also indicates an initial decrease in EF but an improvement after 1 month, with a decrease in LV endsystolic volume. ¹¹ A 4‐month trial of carvedilol similarly produced improvement in ventricular performance parameters. ⁵⁸ When metoprolol was compared with carvedilol in a group of 41 patients with HF and EF <0.26, both groups demonstrated significant but comparable increases in EF. However, carvedilol, but not metoprolol, was associated with a significant reduction in LV enddiastolic volume. ⁵⁹ In a meta‐analysis by Packer et al. ⁶⁰ of four trials (n=248) that compared carvedilol directly to metoprolol, the mean LVEF increased more in the carvedilol groups than in the metoprolol groups (0.084 with carvedilol vs. 0.057; p=0.009) (Figure 2).

Figure 2.

Mean change in left ventricular (LV) ejection fraction with metoprolol and carvedilol therapy in 248 patients enrolled in four trials that directly compared the effects of the two drugs. Adapted with permission from Am Heart J. 2001;141:899‐907.⁶⁰

Clinical and Practical Considerations

Beta blockers are currently recommended, in combination with an ACE inhibitor and a diuretic, for all stable patients with symptomatic HF due to LVSD. Patients should receive both neurohormonal antagonists indefinitely, even if they have responded favorably to diuretics and salt restriction, because these agents have been demonstrated to favorably influence long‐term mortality and morbidity risk in patients with LVSD. The use of ARBs is indicated for those patients who are intolerant of ACE inhibitors or who remain hypertensive after maximal dosage of ACE inhibitors and β blockers. ⁶¹ Diuretics should ordinarily be continued in these patients after euvolemia has been achieved, to prevent recurrent fluid retention. ³ , ⁶² Beta blockers should be withheld in the presence of marked hemodynamic instability or severe decompensation requiring IV inotropes, vasodilators, or high‐dose IV diuretic therapy. They should likewise be used cautiously and at low doses in the presence of asymptomatic bradycardia and usually withheld in patients with bradycardia (<50 bpm) or advanced heart block without a pacemaker. Symptomatic hypotension may be lessened by administering β blockers several hours apart from other drugs with vasodilating actions, especially ACE inhibitors. Signs of fluid retention during dose titration can be treated by increasing diuretic dosage, and symptomatic hypotension may be ameliorated by temporarily reducing ACE inhibitor (or ARB) dosage.

For patients admitted to the hospital with cardiac decompensation, β blockers should be initiated before hospital discharge to improve the likelihood of long‐term compliance, as well as to improve clinical outcomes. ⁴² Importantly, β blockers should not be withheld until patients become refractory to other treatments, since LVSD may progress and patients become exposed to the risk of death during such delays. ³ Beta blockers may be particularly important in preventing sudden cardiac death in patients with HF associated with ischemic heart disease. ⁶³

All β blockers should be initiated at low doses and titrated upwards gradually to achieve the target doses that have been shown to have been effective in the large clinical trials, rather than according to patient therapeutic response. ³ The majority of patients in the β blocker‐HF clinical trials were, in fact, able to tolerate the target dose of the drug being studied; in COMET, ⁵⁰ for example, 75% and 78% of patients achieved their target doses of carvedilol and metoprolol tartrate, respectively. Even in the COPERNICUS ⁴⁰ trial of patients with severe HF, two thirds of patients were able to reach the 25 mg b.i.d. target dose of carvedilol.

CONCLUSIONS

Clinical studies have confirmed the efficacy and safety of β blockers when added to standard therapeutic regimens containing ACE inhibitors and diuretics in improving mortality and morbidity outcomes of patients with HF due to LVSD. Contrary to consensus opinions, recent data have demonstrated that β blockers are safe and provide early clinical benefit when used in HF. Beta‐blocker therapy started in euvolemic patients before hospital discharge not only appears to be safe, but may improve short‐ and long‐term outcomes.

A wide range of evidence demonstrates significant underutilization of β blockers in clinical HF management despite consistent benefits shown in large clinical trials and recommendations for their use in consensus practice guidelines. As with other cardiac drugs, in‐hospital initiation of β‐blocker therapy may improve long‐term utilization by physicians and compliance by patients.

Pharmacologic differences exist among the available β blockers and may influence clinical outcomes in HF treatment. All β blockers cause some degree of acute impairment in cardiac function followed by long‐term improvements in myocardial performance from beneficial effects on ventricular remodeling and contractility. The recent clinical trial that compared metoprolol tartrate and carvedilol found significantly better clinical outcomes with the latter medication.

Acknowledgment and disclosure: This work was supported by GlaxoSmithKline, Philadelphia, PA, with production support by Accel Health, New York, NY.