Deng Jiawen, Heybati Kiyan, Yadav Hemang
Temerty Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada.
Alix School of Medicine, Mayo Clinic, Rochester, Minnesota, USA.
BMJ Open. 2025 Jan 21;15(1):e092594. doi: 10.1136/bmjopen-2024-092594.
Propofol is a widely used sedative-hypnotic agent for critically ill patients requiring invasive mechanical ventilation (IMV). Despite its clinical benefits, propofol is associated with increased risks of hypertriglyceridemia. Early identification of patients at risk for propofol-associated hypertriglyceridemia is crucial for optimising sedation strategies and preventing adverse outcomes. Machine-learning (ML) models offer a promising approach for predicting individualised patient risks of propofol-associated hypertriglyceridemia.
We propose the development of an ML model aimed at predicting the risk of propofol-associated hypertriglyceridemia in ICU patients receiving IMV. The study will use retrospective data from four Mayo Clinic sites. Nested cross validation (CV) will be employed, with a tenfold inner CV loop for model tuning and selection as well as an outer loop using leave-one-site-out CV for external validation. Feature selection will be conducted using Boruta and least absolute shrinkage and selection operator-penalised logistic regression. Data preprocessing steps include missing data imputation, feature scaling and dimensionality reduction techniques. Six ML algorithms will be tuned and evaluated. Bayesian optimisation will be used for hyperparameter selection. Global model explainability will be assessed using permutation importance, and local model explainability will be assessed using SHapley Additive exPlanations.
The proposed ML model aims to provide a reliable and interpretable tool for clinicians to predict the risk of propofol-associated hypertriglyceridemia in ICU patients. The final model will be deployed in a web-based clinical risk calculator. The model development process and performance measures obtained during nested CV will be described in a study publication to be disseminated in a peer-reviewed journal. The proposed study has received ethics approval from the Mayo Clinic Institutional Review Board (IRB #23-0 07 416).
丙泊酚是一种广泛用于需要有创机械通气（IMV）的重症患者的镇静催眠药物。尽管丙泊酚具有临床益处，但它与高甘油三酯血症风险增加有关。早期识别有丙泊酚相关性高甘油三酯血症风险的患者对于优化镇静策略和预防不良后果至关重要。机器学习（ML）模型为预测丙泊酚相关性高甘油三酯血症的个体化患者风险提供了一种有前景的方法。
我们提议开发一种ML模型，旨在预测接受IMV的ICU患者发生丙泊酚相关性高甘油三酯血症的风险。该研究将使用来自梅奥诊所四个地点的回顾性数据。将采用嵌套交叉验证（CV），使用十倍的内部CV循环进行模型调整和选择，以及使用留一站点法CV的外部循环进行外部验证。将使用Boruta以及最小绝对收缩和选择算子惩罚逻辑回归进行特征选择。数据预处理步骤包括缺失数据插补、特征缩放和降维技术。将调整和评估六种ML算法。将使用贝叶斯优化进行超参数选择。将使用排列重要性评估全局模型可解释性，使用SHapley加性解释评估局部模型可解释性。
所提议的ML模型旨在为临床医生提供一种可靠且可解释的工具，以预测ICU患者发生丙泊酚相关性高甘油三酯血症的风险。最终模型将部署在基于网络的临床风险计算器中。在一项将在同行评审期刊上发表的研究报告中，将描述在嵌套CV期间获得的模型开发过程和性能指标。所提议的研究已获得梅奥诊所机构审查委员会的伦理批准（IRB #23 - 0 07 416）。
