毛雨婷 1云洁 2何沁芮 1吴婷婷 1张曼 1程春芳1
作者信息
- 1. 成都中医药大学,四川 成都 610075
- 2. 成都中医药大学附属医院,四川 成都 610032
- 折叠
摘要
Abstract
Objective:To systematically evaluate the risk prediction model for detecting the conversion of mild cognitive impairment(MCI)to Al-zheimer's disease(AD),so as to provide reference for early identification and intervention of MCI.Methods:We searched CNKI,VIP,Wan fang,Chi-nese biomedical literature database,PubMed,EMBASE,Web of Science,Cochrane,and 2023 until March 20,two members independently com-pleted literature screening and quality risk assessment in accordance with the platting criteria.Results:A total of 28 studies were included,and 49 pre-dictive models were constructed using 20 methods,including logistic regression,random forest,etc.18 were high risk,1 was unknown risk,and the conversion rate of dementia was 22.74%～56.56%in MCI conversion patients,the area under the curve(AUC)was(0.692,0.998),47 models were in-ternally validated with cross-validation,2 were not mentioned,7 models were externally validated by authors or other investigators,and 42 models were not externally validated.Conclusion:The prediction models summarized in this system have good prediction performance and good discrimina-tion degree,but some models lack external validation.Future research can further optimize and external validation of existing models,and suggest standardized construction of relevant prediction models for our country's MCI patients at all stages.关键词
轻度认知障碍/阿尔茨海默病/转化/预测模型/动态预测/系统评价Key words
cognitive dysfunction/Alzheimer's disease/prediction model/dynamic prediction/systematic evaluation分类
医药卫生
