Abstract
OBJECTIVE: To investigate the risk factors for postoperative loss of reduction in unstable distal radius fractures and to develop a predictive model, providing clinicians with a more precise risk assessment tool to support the formulation of individualized treatment plans. METHODS: A retrospective analysis was conducted on the clinical data of 209 patients with unstable distal radius fractures who met the selection criteria between January 2018 and December 2023. There were 97 males and 112 females, aged 44-81 years with a mean age of 57.1 years. Univariate analysis was performed to identify factors associated with postoperative loss of reduction. Multivariate logistic regression analysis was then used to screen risk factors and construct a nomogram prediction model. The Hosmer-Lemeshow test was applied to assess model fit, while the area under the receiver operating characteristic (ROC) curve (AUC) was calculated to evaluate the predictive performance. Additionally, decision curve analysis was employed to assess the clinical utility of the model. RESULTS: At 6 months after operation, radiographic evaluation showed loss of reduction in 68 cases. Univariate analysis identified the following as influencing factors for postoperative loss of reduction ( P<0.05): age, fracture displacement, fracture classification, dorsal metaphyseal comminuted fracture, osteoporosis, operation time, and preoperative serum calcium level. Multivariate analysis confirmed the following as risk factors for postoperative loss of reduction ( P<0.05): older age, more severe fracture type (higher AO/OTA classification), presence of fracture displacement, dorsal metaphyseal comminuted fracture, low preoperative serum calcium level, osteoporosis, and prolonged operation time. The nomogram prediction model constructed based on these factors demonstrated high accuracy in assessing the risk of loss of reduction, with an AUC of 0.946 (95% CI: 0.917, 0.975). The calibration curve showed good agreement between predicted and observed probabilities ( χ (2)=4.735, P=0.785). Decision curve analysis indicated that when the predicted risk of postoperative loss of reduction exceeds 0.1, timely intervention can yield substantial net clinical benefit. CONCLUSION: Older age, AO/OTA type C fractures, fracture displacement, dorsal metaphyseal comminuted fracture, prolonged operation time, low preoperative serum calcium level, and comorbid osteoporosis are the main risk factors for postoperative loss of reduction in patients with unstable distal radius fractures. The established predictive nomogram model enables clinicians to more accurately assess the risk of postoperative loss of reduction and provides valuable support for personalized treatment decisions.