Methods and practical applications of clinical prediction model development
10.3760/cma.j.cn112338-20250415-00246
- VernacularTitle:临床预测模型构建方法与应用实践
- Author:
Zichen YE
1
;
Jiahui WANG
;
Qu LU
;
Peng XUE
;
Yu JIANG
Author Information
1. 中国医学科学院北京协和医学院群医学及公共卫生学院，北京 100730
- Publication Type:Journal Article
- Keywords:
Clinical prediction model;
Model development;
Model validation;
Study design;
Performance evaluation
- From:
Chinese Journal of Epidemiology
2025;46(9):1640-1649
- CountryChina
- Language:Chinese
-
Abstract:
Clinical prediction models are statistical tools that incorporate multiple variables to predict the likelihood of specific outcomes, by which the accuracy and efficiency of medical decision-making can be facilitated and patient health outcomes can be improved. However, many current studies face problems, such as model construction and reporting irregularities, as well as questionable reliability, which limit their clinical application of clinical prediction model. Therefore, this study systematically reviews relevant literatures, including publications from journals like BMJ, and outline the steps involved in constructing clinical prediction models based on practical research experience. It also provides an in-depth comparison of commonly used methods during the construction process and proposes a comprehensive guiding framework to help researchers in the field to better understand and master the core concepts and practical skills of clinical prediction models for the purpose of improving their professional capabilities in the development, validation, and application of clinical prediction models.
