Chen Binyan, Zhou Jinghao, Chen Shengzhang, Wang Fei, Liu Ping, Xu Ying, Huang Pan, Cai Fuman
College of Nursing, Wenzhou Medical University, Wenzhou, China.
College of Nursing, AKsu Vocational and Technical College, AKsu, China.
J Clin Nurs. 2025 Aug;34(8):3353-3369. doi: 10.1111/jocn.17860. Epub 2025 Jun 12.
This study was to create an interpretable machine learning model to predict the risk of mortality within 90 days for ICU patients suffering from pressure ulcers.
We retrospectively analysed 1774 ICU pressure ulcer patients from the Medical Information Mart for Intensive Care (MIMIC)-IV database.
We used the LASSO regression and the Boruta algorithm for feature selection. The dataset was split into training and test sets at a 7:3 ratio for constructing machine learning models. We employed logistic regression and nine other machine learning algorithms to build the prediction model. Restricted cubic spline (RCS) was used to analyse the linear relationship between the Braden score and the outcome, whereas the SHAP (Shapley additive explanations) method was applied to visualise the model's characteristics.
This study compared the predictive ability of the Braden Scale with other scoring systems (SOFA, APSIII, Charlson, SAPSII). The results showed that the Braden Scale model had the highest performance, and SHAP analysis indicated that the Braden Scale is an important influencing factor for the risk of 90-day mortality in the ICU. The restricted cubic spline curve demonstrated a significant negative correlation between the Braden Scale and mortality. Subgroup analysis showed no significant interaction effects among subgroups except for age.
The machine learning-enhanced Braden Scale has been developed to forecast the 90-day mortality risk for ICU patients suffering from pressure ulcers, and its efficacy as a clinically reliable tool has been substantiated.
Patients or public members were not directly involved in this study.
本研究旨在创建一个可解释的机器学习模型，以预测患有压疮的重症监护病房（ICU）患者90天内的死亡风险。
我们对重症监护医学信息数据库（MIMIC-IV）中的1774例ICU压疮患者进行了回顾性分析。
我们使用LASSO回归和Boruta算法进行特征选择。数据集按7:3的比例分为训练集和测试集，用于构建机器学习模型。我们采用逻辑回归和其他九种机器学习算法来构建预测模型。使用受限立方样条（RCS）分析Braden评分与结局之间的线性关系，而应用SHAP（Shapley加法解释）方法来可视化模型的特征。
本研究比较了Braden量表与其他评分系统（序贯器官衰竭评估量表、急性生理与慢性健康状况评分系统III、查尔森合并症指数、简化急性生理学评分系统II）的预测能力。结果表明，Braden量表模型具有最高的性能，SHAP分析表明Braden量表是ICU患者90天死亡风险的重要影响因素。受限立方样条曲线显示Braden量表与死亡率之间存在显著的负相关。亚组分析显示，除年龄外，各亚组之间无显著的交互作用。
已开发出机器学习增强的Braden量表，以预测患有压疮的ICU患者90天的死亡风险，并且其作为临床可靠工具的有效性已得到证实。
患者或公众成员未直接参与本研究。
