METHODS: 10 010 high-risk noncardiac surgical patients were randomized aspirin or placebo and clonidine or placebo. Neuraxial block was defined as intraoperative spinal anaesthesia, or thoracic or lumbar epidural anaesthesia. Postoperative epidural analgesia was defined as postoperative epidural local anaesthetic and/or opioid administration. We used logistic regression with weighting using estimated propensity scores.

RESULTS: Neuraxial block was not associated with the primary outcome [7.5% vs 6.5%; odds ratio (OR), 0.89; 95% CI (confidence interval), 0.73-1.08; P=0.24], death (1.0% vs 1.4%; OR, 0.84; 95% CI, 0.53-1.35; P=0.48), myocardial infarction (6.9% vs 5.5%; OR, 0.91; 95% CI, 0.74-1.12; P=0.36) or stroke (0.3% vs 0.4%; OR, 1.05; 95% CI, 0.44-2.49; P=0.91). Neuraxial block was associated with less clinically important hypotension (39% vs 46%; OR, 0.90; 95% CI, 0.81-1.00; P=0.04). Postoperative epidural analgesia was not associated with the primary outcome (11.8% vs 6.2%; OR, 1.48; 95% CI, 0.89-2.48; P=0.13), death (1.3% vs 0.8%; OR, 0.84; 95% CI, 0.35-1.99; P=0.68], myocardial infarction (11.0% vs 5.7%; OR, 1.53; 95% CI, 0.90-2.61; P=0.11], stroke (0.4% vs 0.4%; OR, 0.65; 95% CI, 0.18-2.32; P=0.50] or clinically important hypotension (63% vs 36%; OR, 1.40; 95% CI, 0.95-2.09; P=0.09).

METHODS: The PeriOperative Ischemic Evaluation (POISE)-3 Trial is a large international randomized controlled trial designed to determine if TXA is superior to placebo for the composite outcome of life-threatening, major, and critical organ bleeding, and non-inferior to placebo for the occurrence of major arterial and venous thrombotic events, at 30 days after randomization. Using a partial factorial design, POISE-3 will additionally determine the effect of a hypotension-avoidance strategy versus a hypertension-avoidance strategy on the risk of major cardiovascular events, at 30 days after randomization. The target sample size is 10,000 participants. Patients ≥45 years of age undergoing noncardiac surgery, with or at risk of cardiovascular and bleeding complications, are randomized to receive a TXA 1 g intravenous bolus or matching placebo at the start and at the end of surgery. Patients, health care providers, data collectors, outcome adjudicators, and investigators are blinded to the treatment allocation. Patients on ≥ 1 chronic antihypertensive medication are also randomized to either of the two blood pressure management strategies, which differ in the management of patient antihypertensive medications on the morning of surgery and on the first 2 days after surgery, and in the target mean arterial pressure during surgery. Outcome adjudicators are blinded to the blood pressure treatment allocation. Patients are followed up at 30 days and 1 year after randomization.

DISCUSSION: Bleeding and hypotension in noncardiac surgery are common and have a substantial impact on patient prognosis. The POISE-3 trial will evaluate two interventions to determine their impact on bleeding, cardiovascular complications, and mortality.

METHODS: We conducted a trial involving patients undergoing noncardiac surgery. Patients were randomly assigned to receive tranexamic acid (1-g intravenous bolus) or placebo at the start and end of surgery (reported here) and, with the use of a partial factorial design, a hypotension-avoidance or hypertension-avoidance strategy (not reported here). The primary efficacy outcome was life-threatening bleeding, major bleeding, or bleeding into a critical organ (composite bleeding outcome) at 30 days. The primary safety outcome was myocardial injury after noncardiac surgery, nonhemorrhagic stroke, peripheral arterial thrombosis, or symptomatic proximal venous thromboembolism (composite cardiovascular outcome) at 30 days. To establish the noninferiority of tranexamic acid to placebo for the composite cardiovascular outcome, the upper boundary of the one-sided 97.5% confidence interval for the hazard ratio had to be below 1.125, and the one-sided P value had to be less than 0.025.

RESULTS: A total of 9535 patients underwent randomization. A composite bleeding outcome event occurred in 433 of 4757 patients (9.1%) in the tranexamic acid group and in 561 of 4778 patients (11.7%) in the placebo group (hazard ratio, 0.76; 95% confidence interval [CI], 0.67 to 0.87; absolute difference, -2.6 percentage points; 95% CI, -3.8 to -1.4; two-sided P<0.001 for superiority). A composite cardiovascular outcome event occurred in 649 of 4581 patients (14.2%) in the tranexamic acid group and in 639 of 4601 patients (13.9%) in the placebo group (hazard ratio, 1.02; 95% CI, 0.92 to 1.14; upper boundary of the one-sided 97.5% CI, 1.14; absolute difference, 0.3 percentage points; 95% CI, -1.1 to 1.7; one-sided P = 0.04 for noninferiority).