BACKGROUND: Tranexamic acid (TXA) reduces perioperative blood loss, but comparative effectiveness and safety across administration routes remain uncertain. We evaluated topical, intravenous IV, and oral TXA in unilateral total hip and knee arthroplasty. METHODS: Using Taiwan's national health insurance database (2012-2021), we identified patients undergoing unilateral arthroplasty and evaluated perioperative TXA use, red blood cell transfusion, and adverse events within 60 days after discharge. Conditional logistic regression and six propensity-score-matched comparisons were conducted: topical vs none, IV vs none, oral vs none, topical vs IV, topical+ 1.0g IV vs topical alone, and IV+ 1.0g topical vs IV alone. RESULTS: Compared with no TXA, all routes were associated with lower transfusion risk (topical: RR 0.45, 95%CI 0.41-0.48, p 0.999), and the combined use was not associated with a different risk compared with either route alone. Topical TXA was associated with the greatest reduction in predicted transfusion risk at low dose (- 3,087 per 10,000 -2,886, -3,276, 0→3.0g). Overall, TXA use was not associated with increased venous thromboembolism (VTE), infections, or wound complications. A higher VTE incidence was observed with topical TXA in patients with prior vascular disease without pharmacological prophylaxis, but this was not significant after adjustment for confounders. TXA was not associated with renal injury overall, but high-dose IV (>3.0 g) or oral (>8.0 g) TXA increased predicted risk (~10%) in patients with preexisting renal disease. CONCLUSION: TXA use was associated with substantial transfusion reduction without an overall increase in adverse events. No administration route or combination proved superior. Caution is warranted with higher-dose IV or oral TXA in patients with renal disease, and the association between topical TXA and VTE in high-risk patients merits further investigation.
Hsu et al. (2026) studied this question.