Development and validation of a predictive model for refracture risk in elderly individuals with osteoporotic
Xinling Ma1,2, Yiji Li3, Xiaozhen Lin4
1Youjiang Medical University for Nationalities, No.98, Chengxiang Road, Baise, 533000, China. maxl@ymun.edu.cn.
Background:
The rising incidence of refractures and associated adverse outcomes among individuals with osteoporotic vertebral compression fractures has gained significant attention. Identifying refracture risk is crucial for implementing effective preventive strategies. This study aimed to develop a predictive model for assessing the risk of refracture following an initial vertebral fragility fracture.
Methods:
The study included both men and women, focusing on individuals aged 60 and above. The fractures were diagnosed using X-rays. Participants from three centers were used for model development, while the fourth center was used for validation. The discrimination and calibration of each model was assessed using the validation set. A predictive model for preoperative refracture risk was developed using logistic regression analysis and visualized via a nomogram.
Results:
A total of 1,632 participants with a median age of 72 years were enrolled, of whom 428 had experienced a refracture. Eleven factors were identified as significant predictors of refracture: age, rehabilitation status, treatment regimen, alcohol consumption, body mass index (BMI), duration of medication use, history of comorbidities, fall risk, and levels of bone gla protein (BGP), total alkaline phosphatase (TALP), and homocysteine (Hcy). The average AUC of the nomogram was 0.936 (0.92-0.952) in the training set and 0.922 (0.894-0.949) in the validation set. The values for sensitivity, specificity, and accuracy were 0.825, 0.908, and 0.885, respectively.
Conclusion:
Eleven independent risk factors for refracture were identified and incorporated into a nomogram, which was subsequently converted into an online calculator. This tool can be used to evaluate the risk of refracture in individuals with initial fragility fractures, showing the potential to inform clinical decision-making and optimize patient management. The global applicability of the developed tool still needs to be tested and calibrated by including diverse populations, as this study only included Chinese participants during the tool development phase.
Clinical Registration Number:
Not applicable.
