Suture-Button Fixation Provides a Statistically But Not Clinically Important Improvement in Function for Syndesmotic Injuries in Unstable Ankle Fractures: A Systematic Review and Network Meta-Analysis of Randomized Controlled Trials.

This is the authors' abstract. We don't add key points for this paper.

Background: No consensus exists on the optimal syndesmotic fixation for unstable ankle fractures. We compared strategies using a network meta-analysis (NMA) of randomized controlled trials (RCTs). Methods: Seven databases and trial registers were searched through November 1, 2025 for RCTs of adults undergoing fixation of an ankle fracture with syndesmotic injury; quasirandomized and duplicate/follow-up reports were excluded. Interventions were grouped a priori as suture-button/dynamic fixation, metallic 3-cortex screw (reference), metallic 4-cortex screw, bioabsorbable screw, and no fixation. A frequentist random-effects NMA estimated mean differences (MD) for function (Olerud-Molander Ankle Score [OMAS] at 1 year, the outcome reported by the most trials; AOFAS) and odds ratios (OR) for malreduction and reoperation (primary definition excluding planned implant removal), with treatments ranked by P-score and effects interpreted against the minimal clinically important difference (MCID). Certainty was rated with a CINeMA-equivalent assessment. Results: = 10%) without inconsistency. The AOFAS did not differ between suture-button and the 3-cortex screw (MD 0.0; 95% CI, -4.5 to +4.5). Malreduction was lower with suture-button fixation (OR 0.26; 95% CI, 0.06-1.11; 0.27 [0.11-0.64] fixed-effect). Unplanned reoperation did not differ between constructs (suture-button OR 0.45; 95% CI, 0.17-1.16), whereas all-cause reoperation favored the suture-button and bioabsorbable constructs. Certainty of evidence was low to very low. Conclusions: Suture-button/dynamic fixation provides 1-year function at least equivalent to screws, with lower malreduction and fewer all-cause reoperations from avoiding routine implant removal. Functional differences are small and below the MCID, and certainty is low. Level of Evidence: Therapeutic, Level I. See Instructions for Authors for a complete description of levels of evidence. Registration: PROSPERO CRD420251166994.