Background Second fracture following initial hip fracture surgery poses a life-threatening risk in the elderly population. This study aims to investigate the associated risk factors and establish a prediction model.
Methods Retrospectively data were obtained from two affiliated hospitals at Soochow University for older patients diagnosed with hip fractures who underwent surgical treatment between January 2019 and December 2021. The endpoint was a second fracture. Independent risk factors for second fractures in patients were identified through the least absolute shrinkage and selection operator (LASSO) regression and multivariate Cox regression. A nomogram was established and assessed for predictability, discriminatory ability, and clinical applicability using areas under the receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA) in both cohorts.
Results: Among 1735 patients, 8.7% (115/1735) had second fractures within two years after surgery. Variables screened by LASSO, including age, hip joint function, neurovascular disease, eye disease, living alone, and regular exercise, were incorporated into the Cox regression model. The nomogram demonstrated favorable discriminatory ability, with areas under the ROC curves (AUC) of 0.832 (95% CI, 0.765-0.895) and 0.773 (95% CI, 0.727-0.818) after development and validation, respectively. The calibration curves showed good consistency between the actual second fracture incidence and the predicted probability. DCA of the nomogram demonstrated the model’s excellent clinical efficacy.
Conclusions The nomogram model enabled accurate individualized prediction of second fractures in elderly patients within two years after surgical treatment, which might assist clinicians in precise perioperative management and rehabilitation education following initial hip surgery.