Abstract
<jats:p>Background. Gunshot femoral fractures are associated with a high risk of infectious complications, which significantly worsens treatment outcomes. Early identification of prognostic factors for the development of infectious complications at the time of hospital admission remains a relevant clinical challenge. The purpose was to develop a predictive model and to determine the risk factors for infectious complications in patients with gunshot-related femoral fractures. Materials and methods. A retrospective study of 98 patients with gunshot femoral fractures was conducted. Thirteen potential risk factors were analyzed, including injury characteristics, harmful habits, and anthropometric parameters. Multivariable logistic regression was used to identify independent predictors of infectious complications. The predictive performance of the model was assessed using receiver operating characteristic analysis. Results. At 6 months of follow-up, a positive treatment outcome was achieved in 66 (67.3 %) patients, while the overall risk of infectious complications was 32.7 % (95% CI 23.7–42.3 %). Univariate analysis revealed a significantly increased risk among smokers, patients with soft tissue defects of types β and γ, bone defects of types II–IV, and those with combined injuries. Multivariable analysis identified body mass index, smoking, and the type of bone defect as independent predictors of infectious complications. The three-factor logistic regression model demonstrated good predictive performance (AUC = 0.866; 95% CI 0.787–0.946) with an optimal cutoff value of 0.325, sensitivity of 87.5 %, and specificity of 87.9 %. Conclusions. The combination of local severity of bone tissue injury and modifiable risk factors plays a decisive role in predicting infectious complications in gunshot femoral fractures. Application of the proposed model at the time of hospital admission allows for identification of high-risk patients and treatment optimization.</jats:p>