Skip navigation

Beta blockers and mortality in heart failure

 

Clinical bottom line

Beta blockers reduce the risk of mortality or the need for hospital admission for heart failure by about a third.


Reference

MC Shibata et al. Systematic review of the impact of bet blockers on mortality and hospital admissions in heart failure. European Journal of Heart Failure 2001 3: 351-357.

Review

Trials were sought from databases and previous reviews. For inclusion they compared oral beta blocker with inactive control, had randomised parallel group design, had information on mortality, and on hospital admissions. The date of the last search was January 2000.

Results

There were 22 eligible studies (all randomised and double blind) with 10,480 patients. Eight trials with no deaths could not be analysed. The average age was 62 years, 4% were female, and the average ejection fraction was 26%.

Treatments were bisoprolol (2 trials, 31% of treated patients), bucindolol (3, 2%) carvedilol (8, 23%), metopralol (7, 42%) and nebivodol (2, 0.3%). There was wide range of doses. The smallest trial had 12 patients, the largest nearly 4,000.

Mortality

The effect of beta blockers on mortality is shown in Figure 1 and Table 1. In 14 trials with over 10,000 patients, the overall mortality was reduced by beta blockers; in Figure 1 the points are below the line of equality, showing that events occurred less frequently with beta-blocker than with control. The relative risk of death with beta blocker compared with inactive control was 0.62 (95%CI 0.55 to 0.69). The number needed to treat was 17 (14 to 22). This means that for every 17 patients treated with beta blocker one would not have died who would have with inactive control.

Figure 1: Mortality with beta blocker and control.


Mortality and hospital admission for heart failure

The effect of beta blockers on mortality and hospital admission as a combined outcome is shown in Figure 2 and Table 1. In nine trials with just under 10,000 patients, overall mortality or hospital admission was reduced by beta blockers; in Figure 2the points are below the line of equality, showing that events occurred less frequently with beta-blocker than with control. The relative risk of death or hospital admission with beta blocker compared with inactive control was 0.81 (95%CI 0.76 to 0.86). The number needed to treat was 12 (10 to 16). This means that for every 12 patients treated with beta blocker one would not have died or been admitted to hospital with heart failure who would have with inactive control.

Figure 2: Mortality or hospital admission with beta blocker and control


Table 1: Results with beta blocker and control

Number/Total (%)
Outcome

Number of trials

Beta blocker

Control

Relative risk (95% CI)

NNT (95% CI)

Mortality

14

443/5366 (8)

682/4867 (14)

0.62 (0.55 to 0.69)

17 (14 to 22)

Mortality or hospital admission

9

1401/5035 (28)

1655/4610 (36)

0.81 (0.76 to 0.86)

12 (10-16)

Hospital admission

13

613/5301 (12)

833/4827 (17)

0.67 (0.61 to 0.74)

17 (14 to 23)

Hospital admission

Hospital admission occurred in 613/5301 (11.5% on beta blocker and 833/4827 (17.2%) with inactive control.The number needed to treat to prevent one hospital admission was 17 (14-23).

Comment

This is a rather good review. It makes the point that heart failure is important, and expensive. In 1995, apparently, $3,400 million was paid to Medicare healthcare beneficiaries in the USA for congestive heart failure. Reducing hospital admissions, let alone deaths, is of major importance because of the benefits to costs as well as to patients.