부산대학교 도서관

(1) Background: The Gustilo–Anderson (G/A) grading system is a universally accepted tool used to classify high-grade limb open fractures. The purpose of this study is to find early independent predictors of amputation in emergency settings. (2) Methods: A retrospective analysis involving patients treated at our center between 2010 and 2016 was conducted. Patients with at least one G/A grade III fracture or post-traumatic amputation were included. Three groups were identified: G/A IIIA (A group), G/A IIIB-C (BC group), and Amputation group (AMP group). Each group was further divided into two subgroups considering timing of coverage (early vs. delayed). Univariate and multivariate logistic regression models were developed to identify independent predictors of the limb’s outcome. (3) Results: One-hundred-six patients with G/A III A-B-C fractures or amputation of the affected limb were selected from the Niguarda Hospital Trauma Registry. The patients were divided into the A group (26), BC group (66), and AMP group (14). The rate of infectious complications following early or delayed coverage was evaluated: A group, 9.1% vs. 66.7% (p > 0.05); BC group, 32% vs. 63.6% (p = 0.03); and AMP group, 22% vs. 18.5% (p > 0.05). After further recategorization, the BC subgroups were analyzed: multivariate logistic regression model identified systolic blood pressure (SBP) p = 0.03) and Mangled Extremity Severity Score MESS ≥ 7 (p = 0.001) were determined to be independent predictors of limb amputation. (4) Conclusion: MESS and SBP serve as predictors of amputation. Based on the results, we propose a new management algorithm for mangled extremities. Early coverage is related to lower rates of infectious complications. Referral to high-volume centers with specific expertise is mandatory to guarantee the best results.