Risk prediction models for readmission in patients after percutaneous coronary intervention: a systematic review and critical appraisal
10.3760/cma.j.cn211501-20240508-01143
- VernacularTitle:经皮冠状动脉介入治疗术后患者再入院风险预测模型的系统评价
- Author:
Yanan LI
1
;
Xiujie SUN
;
Wenxin SUN
;
Xiuyan LU
;
Fangyu XIE
Author Information
1. 青岛市市立医院心内一科，青岛 266000
- Publication Type:Journal Article
- Keywords:
Percutaneous coronary intervention;
Systematic review;
Readmission;
Prediction model
- From:
Chinese Journal of Practical Nursing
2025;41(3):197-205
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To systematically evaluate the risk prediction model for readmission in patients after percutaneous coronary intervention, and to provide reference for medical staff to apply or optimize relevant risk assessment tools.Methods:China National Knowledge Infrastructure, VIP, Wanfang Data, China Biomedical Literature Database, and Cochrane Library, PubMed, Embase, CINAHL, Scopus and Web of Science were searched for the related studies on readmission risk prediction models in patients with percutaneous coronary intervention from the establishment of the databases to April 25, 2024. Two researchers independently screened the literature, extracted data, and evaluated the risk of bias in the included studies.Results:A total of eleven articles were included, involving sixteen readmission risk prediction models, with readmission rates ranging from 0.70% to 31.44% and the areas under subjects′working characteristic curves ranging from 0.604 to 0.899. Calibration methods were reported in ten models, five studies reported processing methods of missing data, and external validation was used in three studies. The overall risk of bias was higher. The top six predictors of repeated reports in the readmitted model were age, renal insufficiency, sex, congestive heart failure, diabetes and health insurance.Conclusions:The readmission risk prediction models had good predictive performance. However, the quality of the model methodology was limited. It is necessary to improve the research quality in data sources, measurement and definition of predictive factors, processing of missing data and model evaluation. In the future, data mining can be used to apply the readmission prediction model in the early stage of admission, so as to identify high-risk patients as early as possible and effectively prevent the occurrence of readmission.
