Current issues of ACP Journal Club are published in Annals of Internal Medicine


Bisoprolol improved function but not survival in heart failure

ACP J Club. 1995 May-June;122:62. doi:10.7326/ACPJC-1995-122-3-062

Source Citation

CIBIS Investigators and Committees. A randomized trial of β-blockade in heart failure. The Cardiac Insufficiency Bisoprolol Study (CIBIS). Circulation. 1994 Oct;90:1765-73.



To study whether bisoprolol improves survival in patients with heart failure.


Randomized, placebo-controlled, double-blind trial with a mean follow-up of 1.9 years.


Hospitals in Europe.


641 patients (mean age, 60 y; 83% men) with chronic heart failure (New York Heart Association [NYHA] class III [95%] or IV). Patients had to be age 18 to 75 years, ambulatory, not awaiting cardiac transplantation, taking diuretics and vasodilators, and had to have a left ventricular ejection fraction < 40%. Exclusion criteria were heart failure caused by hypertrophic or restrictive cardiomyopathy with left ventricular diastolic dysfunction or secondary to unrepaired mitral or aortic valve disease; myocardial infarction within 3 months; shortened life expectancy; thyroid dysfunction; insulin-dependent diabetes mellitus; asthma; renal insufficiency; resting heart rate < 65 beats/min; or systolic blood pressure < 100 mm Hg or > 160 mm Hg immediately before randomization.


320 patients were allocated to bisoprolol, 1.25 mg/d, which was doubled at both 48 hours and 1 month. 321 patients were allocated to placebo. Amiodarone and digitalis could be used. Follow-up was > 99%.

Main Outcome Measures

Total mortality and treatment withdrawals. Secondary end points were morbidity and adverse effects.

Main Results

Intention-to-treat analysis was used. The groups did not differ for total mortality (20.9% in the placebo group vs. 16.6% in the bisoprolol group; P = 0.22), cause-specific mortality, mode of death (sudden, related to ventricular tachycardia, or instantaneous), rate of premature withdrawals, and numbers of patients with a decrease in NYHA functional class levels. Compared with patients receiving placebo, patients receiving bisoprolol had fewer episodes of heart failure decompensation requiring hospitalization (61 vs. 90 patients; P < 0.01), fewer nonlethal events (acute pulmonary edema, heart failure without pulmonary edema, and cardiogenic shock) (107 vs. 154 episodes; P < 0.001), fewer cases of documented ventricular tachycardia (3 vs. 11 patients; P = 0.03), and improved functional status (21% of the patients receiving bisoprolol had an increase in at least 1 NYHA functional class level vs. 15% of patients receiving placebo; P = 0.04).


Bisoprolol did not decrease mortality but did increase functional status and reduce nonfatal events in patients with heart failure.

Source of funding: In part, E. Merck Company.

For article reprint: Dr. P. Lechat, Hôpital Pitié-Salpêtrière, Service de Pharmacologie, 47 Boulevard de L'Hôpital, 75013 Paris, France. FAX 33-1-42-161-688.


The theoretical rationale for using β-blockade in patients with congestive heart failure (CHF) to reduce harmful effects of compensatory sympathetic activity has withstood the test of time. Before it can be recommended clinically, however, several questions need to be answered: which β-blockers to use, how long treatment should last, how many patients will not tolerate therapy, and what is the comparative benefit for idiopathic or ischemic cardiomyopathy? Previous studies and this trial of bisoprolol suggest benefit. Although mortality was not reduced, bisoprolol was associated with fewer episodes of decompensated CHF, fewer episodes of ventricular tachycardia, and improved functional status.

Nevertheless, clinicians generally should refrain from "trying this at home" because the trial also shows difficulties in identifying patients likely to benefit. Subgroup analyses suggested a selective benefit in idiopathic compared with ischemic cardiomyopathy and in NYHA class IV compared with class III failure. Only 59% of the bisoprolol group received the intended maximum dose. 40% were receiving unidentified vasodilators other than angiotensin-converting-enzyme inhibitors or calcium antagonists. Most patients (95%) had class III failure, but the investigators suspect that many of these actually may have been class II; 44% of patients were not receiving digitalis.

Only half the patients achieved the target dose, and mortality was lower than the investigators anticipated. These facts reduce the statistical power to show benefit, or they may indicate reduced potential benefit because of positive effects of other current therapies. The authors note that the confidence interval for the nonsignificant risk reduction in mortality (0.56 to 1.15) "remains compatible with a significant reduction in mortality" from bisoprolol. The span of the confidence interval across 1, however, is also compatible with no treatment effect or increased mortality.

Practicing clinicians should be aware of the research question and be prepared for trials that carry β-blockade in CHF from the research arena to clinical use with clear indications.

Valerie A. Lawrence, MD
William D. Linn, PharmD University of Texas San Antonio, Texas