Nine RCTs were included in this review (102,621 participants). Sample sizes ranged from 1,276 to 39,876. Follow-up ranged from 4.4 to 10.1 years.
A significant difference favoured aspirin for major cardiovascular events (RR 0.90, 95% CI 0.85 to 0.96; nine studies). Over a mean follow-up period of 6.9 years this related to an ARR of 0.39% (95% CI 0.18% to 0.61%) and corresponded to a NNT of 253 (95% CI 163 to 568) to prevent one major cardiovascular event.
A significant difference favoured control treatment for haemorrhagic stroke (RR 1.35, 95% CI 1.01 to 1.81; nine studies). Over a mean follow-up period of 6.9 years this related to an absolute risk increase of 0.06% (95% CI 0.003% to 0.126%) and corresponded to a NNH of 1,560 (95% CI 764 to 33,333) to cause one haemorrhagic stroke.
A significant difference favoured control treatment for major bleeding (RR 1.62, 95%CI 1.31 to 2.00; nine studies). Over a mean follow-up period of 6.9 years this related to an absolute risk increase of 0.38% (95% CI 0.21% to 0.55%) and corresponded to a NNH of 261 (95% CI 182 to 476) to cause one major bleeding event.
There was no significant difference in the rate of fatal and nonfatal myocardial infarction, nonfatal myocardial infarction, fatal and nonfatal stroke, nonfatal stroke (eight studies), death, ischaemic stroke and cardiovascular death. For myocardial infarction there was significant heterogeneity of 63%. No significant heterogeneity was found for any other analysis. Meta-regression did not reveal relationships between any of the investigated covariates and the effect of aspiring on major cardiovascular events and major bleeding.
No dose or formulation effects of aspirin on major cardiovascular events were found.
The relative risk of major cardiovascular events was similar across all sensitivity and subgroup analyses.