Home    中文  
 
  • Search
  • lucene Search
  • Citation
  • Fig/Tab
  • Adv Search
Just Accepted  |  Current Issue  |  Archive  |  Featured Articles  |  Most Read  |  Most Download  |  Most Cited

Chinese Journal of Geriatric Orthopaedics and Rehabilitation(Electronic Edition) ›› 2022, Vol. 08 ›› Issue (06): 350-360. doi: 10.3877/cma.j.issn.2096-0263.2022.06.006

• Osteoporosis • Previous Articles     Next Articles

Establishment and Evaluation of Clinical Predictive Model for Osteoporotic fracture

Yan Wang1, Wenjing Li1, Hongzhi Lyu1, Jing Li2, Juan Wang1, Chuan Ren1, Yingze Zhang1,()   

  1. 1. Hebei Orthopaedic Research Institute, Department of Hebei Orthopaedic Biomechanics key Laboratory, Trauma first Aid Center of the Third Hospital of Hebei Medical University, Shijiazhuang 050051, China
    2. Statistics Office of the Third Hospital of Hebei Medical University, Shijiazhuang 050051, China
  • Received:2022-06-30 Online:2022-12-05 Published:2023-01-20
  • Contact: Yingze Zhang
  • About author:
    Wang Yan and Li Wenjing are equally contribute to this work

Abstract:

Objective

To explore the influencing factors of osteoporotic fracture in the elderly and establish a predictive model.

Methods

The fracture patients over 65 years old treated in the third Hospital of Hebei Medical University from January 1, 2021 to December 31, 2021 were collected retrospectively and randomly divided into modeling group and internal verification group at the proportion of 7:3. In addition, fracture patients over 65 years old treated in the third Hospital of Hebei Medical University from January 1, 2022 to March 31, 2022 were collected as external verification group, and the data were obtained through medical record inquiry. Multivariate logistic regression analysis was used to screen the influencing factors of osteoporotic fracture. A clinical prediction model of osteoporotic fracture was constructed in the modeling population, which was verified as a whole, internally and externally. The differentiation, calibration and clinical effectiveness of the model were evaluated, and the corresponding diagrams were drawn.

Results

A total of 2, 512 patients with fracture met the inclusion exclusion criteria, including 936 males (37.3%) and 1, 576 females (62.7%). They were divided into two groups: development group (n=1, 751) and internal validation group (n=761). A total of 599 patients with fracture were included in the external validation group. According to multivariate logistic regression analysis, female (OR=1.944, 95% CI: 1.569, 2.409)、age≥75 years old (75-84 years old, OR=2.150, 95% CI: 1.718, 2.714; 85-94 years old, OR=4.285, 95% CI: 2.936, 6.198) 、BMI<18.5 kg/m2 (OR=1.885, 95% CI: 1.180, 3.087) 、hypertension (OR=1.441, 95% CI: 1.168, 1.779) and hypoproteinemia (OR=2.484, 95% CI: 1.987, 3.113) were independent risk factors for osteoporotic fracture, which were included in the predictive model and evaluated. The AUC value of the model was 0.734 (95% CI: 0.711, 0.757), 0.717 (95% CI: 0.681, 0.734) in the internal verification group and 0.717 (95% CI: 0.656, 0.759) in the external verification group. The calibration maps verified by the development group, internal and external validation group all show that the prediction ability of the model is excellent, and the P value of H-L test are all higher than 0.05. The DCA curve indicated that the clinical validity of the model was the best when the threshold probability was 0.3-0.8.

Conclusions

Women, over 75 years old, BMI<18.5 kg/m2, people with hypertension and hypoproteinemia are prone to osteoporotic fracture. The clinical prediction model established in this study can better predict the risk of osteoporotic fracture.

Key words: Osteoporosis, Fracture, Clinical predictive model, Nomogram

京ICP备07035254号-18
Copyright © Chinese Journal of Geriatric Orthopaedics and Rehabilitation(Electronic Edition), All Rights Reserved.
Tel: 0311-88603818 E-mail: zhlngkykf@126.com
Powered by Beijing Magtech Co. Ltd