中医临床预测模型研究的概况性综述
Scoping review of study on clinical prediction model of traditional Chinese medicine
景城阳 1冯琳 2李嘉琛 2梁立荣 2廖星1
作者信息
- 1. 中国中医科学院中医临床基础医学研究所,北京 100700
- 2. 首都医科大学附属北京朝阳医院北京市呼吸疾病研究所,北京 100020
- 折叠
摘要
目的:基于概况性综述的知识综合方法,分析和总结中医临床预测模型研究现况.方法:计算机检索7个中英文数据库,检索时段均为自建库至2023年3月31日.结果:共纳入113项中医临床预测模型研究,涵盖疾病数量48个.建模数据最多来源于横断面研究,共包含12类预测因子,预测结局最多者为中医证型.21种建模方法中最常用者为Logistic回归,14种区分度指标中最常用者为准确率.仅有23项和6项研究分别评价了模型的校准度和临床效益.75项研究进行了内部验证;仅有6项研究进行了外部验证.结论:中医临床预测模型的应用范围已涵盖多个疾病领域,并呈现出鲜明的中医药特色.然而现有研究在模型的数据来源、评价维度、验证过程和报告规范方面尚存在不足.
Abstract
Objective:Based on the knowledge synthesis methods of scoping review,to analyze and summarize the current status of studies on clinical prediction model of traditional Chinese medicine(TCM).Methods:The seven Chinese and English databases were electronically searched from inception to March 31,2023.Results:A total of 113 studies published were included,covering a total of 48 conditions.Most of the modeling data come from cross-sectional studies,while there were 12 types of predictive factors,and the most prediction type was TCM syndrome.A total of 21 modeling methods were used in all studies,of which the most commonly used was Logistic regression.A total of 14 discrimination evaluation indicators were used,of which the most commonly used was the accuracy.Only 23 studies evaluated the calibration of the models,while only 6 studies evaluated the clinical benefit of the models through decision curve analysis.A total of 75 studies conducted internal validation on model performance as well as only 6 studies conducted external validation.Conclusion:The application of clinical prediction model of TCM has covered a variety of population and disease fields with distinct characteristics of TCM.However,some deficiencies in the data source,evaluation dimension,validation process and reporting specification are still existing in present studies.
关键词
临床预测模型/诊断模型/预后模型/应用领域/模型特征/中医/概况性综述Key words
Clinical prediction model/Diagnostic model/Prognostic model/Application field/Model characteristics/Ttraditional Chinese medicine/Scoping review引用本文复制引用
出版年
2024
NETL
NSTL
万方数据