Identification of risk factors and development of a high-performance predictive model for non-healing in elderly patients with intertrochanteric fractures post-internal fixation.

Abstract

Objective: To identify risk factors associated with non-healing in elderly patients with intertrochanteric femoral fractures treated with internal fixation and to develop a predictive model for non-union risk.

Methods: We conducted a retrospective analysis of 889 elderly patients treated with internal fixation for intertrochanteric fractures at Wuxi Xishan People's Hospital from March 2021 to December 2024. Patients were classified into healing (n=806) and poor healing groups (n=83) based on radiographic evidence three months post-surgery. Univariate and multivariate logistic regression analyses were used to identify significant risk factors. A predictive model was developed and validated using receiver operating characteristic (ROC) analysis and the area under the curve (AUC).

Results: Significant risk factors for poor healing included smoking history (Odds ratio [OR] 1.750, P=0.022), osteoporosis (OR 2.055, P=0.003), posterior or medial wall bone defects (OR 1.964, P=0.005), low postoperative albumin (OR 1.674, P=0.032), and early weight-bearing (OR 1.765, P=0.018). The use of proximal femoral nail antirotation (PFNA) significantly reduced the risk of poor-healing (OR 0.515, P=0.006). The combined predictive model achieved an AUC of 0.949, indicating high predictive value.

Conclusions: Our findings highlight key risk factors for non-healing in elderly patients post-internal fixation for intertrochanteric fractures. The developed predictive model, incorporating clinical, biochemical, and surgical factors, offers high accuracy and may help identify high-risk patients for targeted intervention.