A hospital policy of tranexamic acid to reduce transfusion in major non-cardiac surgery: The traction trial

Abstract BACKGROUND Whether a hospital policy of tranexamic acid administration for patients undergoing major noncardiac surgery safely reduces red blood cell (RBC) transfusion without increasing thrombotic risk is uncertain. METHODS We conducted a multicenter, registry-based, randomized, placebo-controlled, cluster-crossover trial enrolling patients undergoing noncardiac surgeries at high risk for RBC transfusion. Hospitals were randomized at 4-week intervals to a hospital-wide policy of intraoperative tranexamic acid or placebo. Co-primary effectiveness and safety outcomes were the proportion of patients transfused RBCs during hospital admission, and the proportion of patients diagnosed with venous thromboembolism within 90 days. The safety of the tranexamic policy compared to placebo was based on a non-inferiority margin where the upper boundary of the two-sided 95% confidence interval for the odds ratio of venous thromboembolism had to be below 1.47. Secondary outcomes included the number of RBC units transfused, in-hospital diagnoses of myocardial infarction, stroke, deep vein thrombosis or pulmonary embolism, hospital length of stay, intensive care unit admission, hospital survival and overall survival to day 90. RESULTS The co-primary outcomes were evaluable for 8273 patients enrolled across 10 Canadian hospitals. Baseline characteristics were similar between the two groups. The most common surgical specialties included general surgery (n=2742; 33.1%), gynecology (n=1540; 18.6%), urology (n=1434; 17.3%), vascular surgery (n=582; 7.0%), and spine surgery (n=461; 5.6%). Most of the population underwent surgery for an oncologic indication (n=5002; 60.5%). The proportion of patients transfused RBCs during hospital admission was 7.4% (306/4156) in the tranexamic acid group and 9.8% (403/4117) in the placebo group (odds ratio, 0.69; 95% confidence interval CI, 0.54 to 0.89; absolute difference, -2.4 percentage points; 95% CI, -0.8 to -4.6%). The proportion of patients diagnosed with venous thromboembolism was 2.1% (86/4156) in the tranexamic acid group and 2.2% (90/4117) in the placebo group (odds radio, 0.96; 95% CI, 0.64 to 1.42). Analyses used mixed effects models to account for the cluster-crossover design. Patients in the tranexamic acid group received less RBC units compared to those in the placebo group (mean 2.5 units (± 1.43 units) vs. 0.34 units (± 1.86 units) (p0.01). There was no difference in in-hospital diagnoses of myocardial infarction, stroke, deep vein thrombosis or pulmonary embolism. Similarly, there was no difference in hospital length of stay, intensive care unit admission, hospital survival or 90 day survival between the tranexamic acid and placebo group. CONCLUSIONS In patients undergoing major noncardiac surgery, a hospital policy of tranexamic acid administration reduced red cell transfusion without increasing venous thromboembolism. (Funded by the Canadian Institutes of Health Research and others. TRACTION ClinicalTrials.gov number, NCT04803747.