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


Atenolol reduced mortality and cardiovascular events after noncardiac surgery

ACP J Club. 1997 May-Jun;126:58. doi:10.7326/ACPJC-1997-126-3-058

Source Citation

Mangano DT, Layug EL, Wallace A, Tateo I, for the Multicenter Study of Perioperative Ischemia Research Group. Effect of atenolol on mortality and cardiovascular morbidity after noncardiac surgery. N Engl J Med. 1996 Dec 5;335:1713-20.



To determine whether atenolol reduces mortality and cardiovascular events in patients who were discharged after noncardiac surgery and had, or were at risk for, coronary artery disease (CAD).


Randomized, double-blind, placebo-controlled trial.


A U.S. Veterans Affairs medical center.


200 adults (mean age 68 y) hospitalized for elective noncardiac surgery that required general anesthesia and who had either CAD (previous myocardial infarction, typical angina, or atypical angina with a positive stress test) or ≥ 2 risk factors for CAD (age ≥ 65 years, hypertension, smoking, cholesterol level ≥ 6.2 mmol/L, or diabetes mellitus).


Patients received intravenous (IV) placebo (n = 101) or atenolol, 5 mg over 5 minutes (n = 99), 30 minutes before surgery if their heart rate was ≥ 55 beats/min; systolic blood pressure was ≥ 100 mm Hg; and congestive heart failure, third-degree heart block, or bronchospasm was not evident. If these conditions persisted, a second IV dose was given 5 minutes later. After surgery, the drug was given again in the same way. Beginning on the first day after surgery, atenolol was given IV, 5 mg twice daily, or orally if possible, 50 to 100 mg once daily, until hospital discharge (maximum 7 d).

Main outcome measures

Total mortality from discharge to 2 years. The secondary outcome was a combination of myocardial infarction, unstable angina, or congestive heart failure that required hospitalization; myocardial revascularization; or death after discharge.

Main results

Between discharge and 2 years, 9 patients who received atenolol had died (4 from cardiac causes) compared with 21 deaths (12 from cardiac causes) in patients who received placebo (10% vs 21%, P = 0.02). {This absolute risk reduction (ARR) of 11% means that 9 patients would need to be treated (NNT) with atenolol at the time of noncardiac surgery compared with placebo to prevent 1 additional death (95% CI 5 to 76); the relative risk reduction (RRR) was 54%, CI 8% to 78%.}* The groups did not differ for cardiac deaths after discharge (4% vs 12%, P = 0.09), but cardiac events after discharge were lower in the atenolol group {17% vs 32%, P = 0.03; ARR 15%; NNT 7, CI 4 to 34; RRR 47%, CI 11% to 69%}*. For the time from surgery to 2 years, the groups did not differ for total mortality (13% vs 23%, P = 0.07) but cardiovascular mortality was lower in the atenolol group (5% vs 14%, P < 0.05; ARR 9%; NNT 11, CI 6 to 135; RRR 63%, CI 7% to 86%).


For patients who were discharged after elective noncardiac surgery and who had or were at risk for CAD, perioperative atenolol reduced the risk for death and coronary events.

Source of funding: Ischemia Research and Education Foundation.

For article reprint: Dr. D.T. Mangano, San Francisco Veterans Affairs Medical Center, 4150 Clement Street, 129, San Francisco, CA 94121, USA. FAX 415-750-6653.

*Numbers calculated from data in article.


This study by Mangano and colleagues reports that atenolol improved survival at 2 years, although in an earlier abstract no differences were reported in events immediately after surgery (including death, myocardial infarction, unstable angina, congestive heart failure, and ventricular tachycardia) among patients in the atenolol and placebo groups (1).

Mangano and colleagues and others have shown that patients who have postoperative cardiac events (myocardial infarction, ischemia, or unstable angina) have a 28-fold increase in the rate of subsequent cardiac complications during the first 6 months of follow-up (2). Based on these studies, a link should exist between the perioperative complications and long-term outcomes. The principal effect should have occurred in the perioperative period with differences in the rate of cardiac complications, especially myocardial ischemia, immediately after surgery.

Why would 1 week of atenolol make a difference in 2-year mortality if it had no immediate effect after surgery? A potential explanation for this would be that a disproportionate number of patients in the atenolol group continued using β-blockers after discharge, but this was not the case. More patients in the atenolol group seem to have been taking angiotensin-converting enzyme inhibitors, which have been shown to have long-term protective effects.

This study provides another piece of evidence that perioperative β-blockers may be beneficial, but the study would be more convincing if it had shown a link between immediate postoperative outcomes and 2-year mortality.

Mary Charlson, MD
Cornell Medical CenterNew York, New York, USA


1. Wallace A, Layug E, Browner W, et al. Anesthesiology. 1994;81 (Suppl):A99.

2. Mangano DT, Browner WS, Hollenberg M, Li J, Tateo IM. JAMA. 1992;268:233-9.