Guo Pengfei, Wang Lingjie, Gao Zheng, Wang Bin, Yan Wenlong, Qu Xingchi, Yang Sumin
Department of Cardiothoracic surgery, Beijing Shijitan Hospital, Capital Medical University, Beijing, 100038, China.
Department of Cardiothoracic surgery, Affiliated Wenling Hospital, The First People's Hospital of Wenling, Wenzhou Medical University, Zhejiang, 317500, China.
BMC Cardiovasc Disord. 2025 Aug 12;25(1):600. doi: 10.1186/s12872-025-05067-y.
Timely assessment of Low cardiac output syndrome (LCOS) risk after off-pump coronary artery bypass grafting (OPCAB) is crucial, yet hindered by the lack of standardized diagnostic criteria beyond symptoms, therapy response, and ultrasound.
This study aims to develop, construct, and internally validate a predictive model to predict low cardiac output syndrome (LCOS) in patients undergoing off-pump coronary artery bypass grafting (OPCAB).
Using a clinical dataset of 765 OPCAB patients treated between May 2018 and July 2020, encompassing admission, surgical, and postoperative data, a predictive model was developed. Feature selection was performed using Least Absolute Shrinkage and Selection Operator (LASSO) regression. The selected features were then incorporated into a multivariate logistic regression model to establish the final predictor. Model performance was evaluated using the C-index (discrimination), calibration plots (calibration), and decision curve analysis (clinical validity). Internal validation via bootstrap resampling assessed model robustness.
The final prediction model incorporated the following predictors: age, smoking history, ejection fraction, left ventricular end-diastolic diameter, lactic acid levels, room-air pre-operative PaO₂, room-air pre-operative oxygen saturation, carotid artery stenosis, myocardial enzyme levels, internal mammary artery condition, intra-operative blood transfusion volume, intra-operative blood loss, ICU stay duration, ventilator time, and IABP implantation. The model demonstrated excellent discrimination, with a C-index of 0.943 (95% CI: 0.893-0.993) in the derivation cohort and 0.9429 in the bootstrap-validated cohort. Decision curve analysis indicated significant net clinical benefit for model-guided intervention when the predicted LCOS risk threshold exceeded 1%.
This study developed a predictive model that integrates significant risk factors for post-OPCAB LCOS, encompassing demographics, cardiac metrics, laboratory values, procedural variables, and post-operative course. By enabling accurate risk stratification (C-index 0.94), this model may serve as a practical clinical tool to guide timely interventions for high-risk patients.
非体外循环冠状动脉搭桥术（OPCAB）后及时评估低心排血量综合征（LCOS）风险至关重要，但目前除症状、治疗反应及超声检查外，缺乏标准化诊断标准，这一评估受到阻碍。
本研究旨在开发、构建并进行内部验证一个预测模型，以预测接受非体外循环冠状动脉搭桥术（OPCAB）患者的低心排血量综合征（LCOS）。
利用2018年5月至2020年7月期间接受治疗的765例OPCAB患者的临床数据集，该数据集涵盖入院、手术及术后数据，开发了一个预测模型。使用最小绝对收缩与选择算子（LASSO）回归进行特征选择。然后将所选特征纳入多变量逻辑回归模型以建立最终预测指标。使用C指数（辨别力）、校准图（校准）及决策曲线分析（临床有效性）评估模型性能。通过自举重采样进行内部验证以评估模型稳健性。
最终预测模型纳入了以下预测指标：年龄、吸烟史、射血分数、左心室舒张末期直径、乳酸水平、术前室内空气下的PaO₂、术前室内空气下的氧饱和度、颈动脉狭窄、心肌酶水平、乳内动脉状况、术中输血量、术中失血量、重症监护病房停留时间、呼吸机使用时间及主动脉内球囊反搏（IABP）植入情况。该模型显示出优异的辨别力，在推导队列中的C指数为0.943（95%置信区间：0.893 - 0.993），在自举验证队列中为0.9429。决策曲线分析表明，当预测的LCOS风险阈值超过1%时，模型指导的干预具有显著的净临床获益。
本研究开发了一个预测模型，该模型整合了OPCAB术后LCOS的重要风险因素，包括人口统计学特征、心脏指标、实验室值、手术变量及术后病程。通过实现准确的风险分层（C指数为0.94），该模型可作为一种实用的临床工具，用于指导对高危患者进行及时干预。
