Fig. 6

a Subgroup analysis according to the type of surgery. b Trial sequential analysis for outcome of postoperative delirium in patients undergoing cardiac surgery: α = 5% (two-sided) and β = 20% (power, 80%), incidence in control arm calculated from studies = 8.59%, relative risk reduction = 20%. c Trial sequential analysis for outcome of postoperative delirium in patients undergoing major non-cardiac surgery: α = 5% (two-sided) and β = 20% (power, 80%), incidence in control arm calculated from studies 26.44%, relative risk reduction = 20%