Thirty-eight clinical studies were included: 14 were randomised controlled trials (RCTs), three were controlled clinical trials, six were comparative studies, three were multicentre studies, five were case series, one was case control, two were prospective, one was retrospective, and three were case reports.
Continued oral anticoagulant: There were 18 studies (2,155 patients). Four RCTs reported statistically significant superiority of a local haemastatic agent, over another one, for postoperative bleeding events. Three RCTs reported no statistically significant difference in bleeding events.
Continued versus reduced anticoagulant: There were two RCTs and one comparative study (208 patients). All three studies demonstrated no statistically significant difference in postoperative bleeding complications.
Continued versus stopped anticoagulant: There were three RCTs and three studies with other study designs (651 patients). The three RCTs showed no statistically significant difference in bleeding status.
Heparin and reduced or stopped oral anticoagulant: There were two RCTs, and seven studies with other designs (2,514 patients). The two RCTs compared heparin with continued anticoagulation and reported no statistically significant difference in haemorrhage risk.
One study reported that two postoperative thromboembolic events occurred in seven patients who withdrew from oral anticoagulant therapy without bridging therapy (heparin). No other thromboembolic events occurred.
The results of studies with other designs were presented in tables, but not discussed further.