Construction and validation of a risk prediction model for low fall alertness in elderly inpatients
10.16753/j.cnki.1008-2344.2025.01.003
- VernacularTitle:老年住院患者跌倒警觉度低下风险预测模型的构建及验证
- Author:
Xinxin LI
1
;
Xiaoju TENG
;
Xinkai ZHOU
;
Hongmei MA
;
Yating HAN
;
Yingxia LI
;
Jiamei ZHU
;
Kun LUO
Author Information
1. 皖南医学院护理学院,安徽 芜湖 241002
- Publication Type:Journal Article
- Keywords:
elderly inpatients;
fall awareness;
prediction model;
risk factors
- From:
Journal of Shenyang Medical College
2025;27(1):12-19
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To analyze the influencing factors of low fall alertness in elderly inpatients,construct a risk prediction model and validate it,providing a reference for clinical medical staff to identify elderly inpatients with low fall alertness in the early stage.Methods:A total of 605 elderly inpatients treated in Yijishan Hospital affiliated to Wannan Medical College from Oct 2023 to Mar 2024 were enrolled and randomly divided into the training group(n=423)and validation group(n=182)at a ratio of 7∶3.The patients were evaluated using a general information questionnaire,the Social Frailty Screening Tool(HALFT),the Tilburg Frailty Indicator(TFI),and the Self-Awareness of Falls in Elderly scale(SAFE).Multivariate logistic analysis was used to determine the influencing factors of low fall alertness in elderly inpatients.RStudio was used to construct a risk prediction model of low fall alertness.The discrimination,calibration,and clinical net benefit of the model were verified using the receiver operating characteristic(ROC)curves,calibration plots,and decision curve analysis(DCA).Results:Multivariate logistic analysis showed that the history of falls,monthly income,previous physical activity time,social frailty score and TFI score were independent risk factors for low fall alertness in elderly inpatients.The Hosmer-Lemeshow χ2 test showed that χ2=8.863,P=0.354,indicating good calibration of the prediction model.The area under the ROC curve of the training group and the validation group were 0.860(95%CI:0.815-0.904)and 0.937(95%CI:0.888-0.986),respectively,and the maximum Youden indices of the model was 0.576 and 0.788,respectively,indicating good discrimination of the model.The DCA decision curve showed that the model had good clinical effectiveness.Conclusion:The constructed model has a good prediction effect and can help clinical medical staff quickly and effectively screen out elderly inpatients at risk of low fall alertness.
