A total of 33 RCTs was included (n=34,622, range 50 to 3,991 patients). Ten studies were rated high quality on the Jadad scale.
Acute coronary syndrome: β1+2 blockers showed a statistically significant reduction over placebo with respect to all-cause mortality (RR 0.73, 95% CI 0.64 to 0.82; six studies) and vascular events (RR 0.71, 95% CI 0.59 to 0.84; six studies).
There were no statistically significant differences between β1 blockers and placebo in all-cause mortality (five studies) and vascular events (three studies). There was no difference in all-cause mortality rates in direct comparisons between β1 and β1+2 blockers (one study).
Heart failure: β1 blockers in comparison with placebo showed significant reduction in mortality (RR 0.76, 95% CI 0.68 to 0.87; seven studies) but not in vascular events (RR 1.33, 95% CI 0.86 to 2.04; three studies). β1+2 blockers when compared with placebo reduced mortality rates significantly (RR 0.75, 95% CI 0.61 to 0.92; 10 studies) and vascular effects non-significantly (RR 0.80, 95% CI 0.64 to 1.00; six studies).
In direct comparison between β-blockers, all-cause mortality rates were significantly reduced with β1+2 blockers (RR 0.86 95% CI 0.78 to 0.94; four studies) compared with β1 blocking agents.
Statistical heterogeneity was not found to be present in all outcomes except for vascular events when comparing β1+2 blockers with placebo in patients with heart failure (p=0.03, I2=50.1%).
There was funnel plot asymmetry for studies in heart failure, which indicated evidence for publication bias. There was no evidence for publication bias in studies with acute coronary syndrome (funnels plots not presented).
Inclusion of only high-quality trials, analysis of different β-blockers and removal of trials one by one did not seem to influence results. Adjusting for studies with longer follow-up (at least 12 months) did not seem to alter the results significantly.