Influence of Perioperative Antithrombic Agent Discontinuation in Elective Posterior Spinal Surgery: A Propensity-Score-Matched Analysis.

OBJECTIVE

To investigate the influence of perioperative antithrombotic agent (antiplatelet agents and anticoagulants) discontinuation in elective posterior spinal surgery in terms of bleeding complications, such as epidural hematoma and postoperative thromboembolism.

METHODS

We enrolled patients undergoing elective posterior spinal surgery at 9 hospitals between April 2017 and August 2020. We collected data regarding patient baseline characteristics, surgical details, intraoperative estimated blood loss, and postoperative complication rates, including epidural hematoma and thromboembolism. We divided the patients into a discontinuation group, in which antithrombic agents were discontinued perioperatively, and a control group without antithrombic agents. Propensity scores for taking any antithrombic agents were calculated, with 1-to-1 matching based on the estimated propensity scores to adjust for patient baseline characteristics and surgical details. Intraoperative estimated blood loss and 30-day postoperative complication rates were compared between the groups.

RESULTS

We enrolled 9853 patients, including 1123 patients (11.4%) who discontinued antithrombic agents perioperatively. One-to-one propensity score matching yielded 1111 pairs with and without antithrombic agents. Intraoperative estimated blood loss per 10 minutes (8.2 mL vs. 8.9 mL) and the incidence of epidural hematoma requiring revision (0.97% vs. 0.72%) were similar between the groups. Although postoperative cardiac events and stroke were observed only in the discontinuation group (0.27% and 0.09%, respectively), these incidences were not significantly different between the groups.

CONCLUSIONS

Perioperative antithrombic agent discontinuation in elective posterior spinal surgery normalized the intraoperative bleeding tendency and the incidence of postoperative epidural hematoma and did not influence in a significative way the incidence of postoperative thromboembolism.