Seven RCTs met the inclusion criteria (31,286 participants, range 1,279 to 15,526). Follow-up ranged from 26 to 124 weeks. The rate of trial drug discontinuation ranged from 15% to 44%.
New generation oral anticoagulant agents were associated with a three-fold increase in major bleeding events (OR 3.03, 95% CI, 2.20 to 4.16; Ι²=0%) and a more than two-fold increase in any bleeding event (OR 2.26, 95% CI, 2.01 to 2.56; Ι²=8%). There were statistically significant reductions in the risk of stent thrombosis (OR 0.73, 95% CI, 0.54 to 0.98; Ι²=17%) and the composite of ischaemic events (OR 0.86, 95% CI, 0.79 to 0.94; Ι²=0%) but not overall mortality. There was no statistically significant net clinical benefit of new generation oral anticoagulant agents over placebo (OR 0.98, 95% CI, 0.90 to 1.06).
Results from subgroup and sensitivity analyses were reported. There was no evidence of publication bias.