Machine learning-based risk prediction models for acute kidney injury in patients with acute coronary syndrome: A systematic review
- VernacularTitle:基于机器学习的急性冠状动脉综合征患者急性肾损伤风险预测模型的系统评价
- Author:
Qi ZHANG
1
,
2
;
Chenming LI
2
;
Guyue YAN
2
;
Qing WU
1
Author Information
1. Department of Nursing, The First Affiliated Hospital of Soochow University, Suzhou, 215006, Jiangsu, P. R. China
2. School of Nursing, Suzhou Medical College, Soochow University, Suzhou, 215006, Jiangsu, P. R. China
- Publication Type:Journal Article
- Keywords:
Machine learning;
acute coronary syndrome;
acute kidney injury;
risk prediction model;
systematic review
- From:
Chinese Journal of Clinical Thoracic and Cardiovascular Surgery
2026;33(07):1111-1118
- CountryChina
- Language:Chinese
-
Abstract:
Objective To systematically evaluate the risk prediction models for acute kidney injury in patients with acute coronary syndrome (ACS) based on machine learning, providing a reference for clinical selection of appropriate risk assessment tools. Methods Clinical studies using machine learning methods for predicting the risk of acute kidney injury in ACS patients were retrieved from PubMed, Cochrane Library, Embase, Web of Science core database, CNKI, Wanfang Database, CBM, and VIP. The retrieval time was from the establishment of the database to May 24, 2025. The quality of the models were evaluated using the prediction model risk of bias assessment tool. Results Nine articles were included, and a total of 58 prediction models were constructed using 20 machine learning methods. The area under the receiver operating characteristic curve ranged from 0.733 to 0.894. The most commonly used predictors were age and creatinine. The overall bias risk of the included studies was relatively high, but the applicability was good.Conclusion Machine learning models can identify the risk of acute kidney injury in ACS patients. All models have good predictive potential, but they are still in the development stage. It is recommended that future studies adopt prospective design with external validation to improve the stability and predictive accuracy of the models.
