Abstract / Summary
Rhinoplasty is frequently complicated by intraoperative bleeding and early postoperative edema/ecchymosis. Antifibrinolytic (tranexamic acid, TXA) and anti-inflammatory strategies are widely used, but their comparative effectiveness remains uncertain. We aimed to quantify the effects of TXA and systemic corticosteroids versus placebo and versus each other on bleeding, edema, ecchymosis, and surgeon satisfaction in rhinoplasty. We registered a protocol (PROSPERO CRD420251170943) and searched PubMed/MEDLINE, Embase, Scopus, Web of Science, and Cochrane to 14 July 2025 without language limits. Adult rhinoplasty (± septoplasty) randomized trials comparing TXA or systemic corticosteroids with placebo or each other were included. Random-effects models produced mean differences (MD) with 95% CIs; heterogeneity was summarized with I². Predefined time windows prioritized the latest early postoperative assessment. Of 612 records, 18 RCTs ( n = 600) met criteria. Versus placebo, TXA reduced bleeding (MD − 0.82, 95% CI − 0.91 to − 0.73; I²=0%), ecchymosis (MD − 0.79, − 0.88 to − 0.70; I²=0%), and edema (MD − 0.74, − 0.84 to − 0.65; I²=0%). Steroids similarly improved ecchymosis (MD − 0.86, − 0.98 to − 0.74; I²=0%) and edema (MD − 0.75, − 0.85 to − 0.64; I²=0%) and reduced bleeding (MD − 0.71, − 0.82 to − 0.61; I²=0%). In head-to-head analyses, TXA outperformed steroids for ecchymosis (MD − 0.35, − 0.51 to − 0.19; I²=0%) and edema (MD − 0.30, − 0.50 to − 0.11; I²=33%), and increased surgeon satisfaction (MD + 0.40, 0.21 to 0.60; I²=33%). Funnel plots were largely symmetrical for bleeding, ecchymosis, and edema. Risk of bias was generally low to moderate, with occasional unblinded outcome assessment. Both TXA and systemic corticosteroids improve early recovery after rhinoplasty versus placebo; TXA shows a modest edge for soft-tissue outcomes and surgical-field quality. Standardized, blinded, factorial RCTs are needed to define optimal dosing, routes, and any additive effects. Level I, risk / prognostic study.