体外循环下冠状动脉旁路移植术后患者机械通气时间延长风险预测模型的构建与验证 [中文引用][英文引用]

目的  建立体外循环下冠状动脉旁路移植术后患者机械通气时间延长（prolonged mechanical ventilation，PMV）风险预测模型。方法  采用便利抽样法，选取 2021 年 1 月— 2023 年 12 月在北京某三级甲等心血管专科医院接受体外循环冠状动脉旁路移植手术后 2 334 例患者为研究对象。采用结构化语言查询医院电子病历系统，收集研究对象术前、术中和术后资料。按照 3∶1 将研究对象随机分成训练组（n=1 633）和验证组（n=701）。基于训练组数据使用 Logistic 回归构建风险预测模型。采用 Hosmer-Lemeshow 判断模型拟合度，采用受试者工作特征（receiver operating characteristic，ROC）曲线下面积（area undercurve，AUC）检验模型的预测效果。结果  2 334 例患者中 215 例（9.2%）发生机械通气时间延长。基于训练组建立的模型纳入了年龄（OR=1.03）、体质量指数（body mass index，BMI；OR=1.14）、术中体外循环转机时间（OR=1.01）、术中输血（OR=4.15）、术后血清总胆红素（OR=1.08）、术后血清白蛋白（OR=0.92）和术后再次开胸探查（OR=5.49）7 个机械通气时间延长的影响因素。体外循环冠状动脉旁路移植术后患者机械通气时间延长预测模型的 AUC 为 0.761，95%CI（0.716~0.806），最大约登指数为 0.105，灵敏度为 77.94%，特异度为 64.38%。基于验证组数据对模型进行验证，模型 AUC 为 0.733，95%CI（0.662~0.804），灵敏度为 75.32%，特异度为 57.97%，预测正确率为 73.61%。结论  本研究建立的风险预测模型预测效果较好，可为临床护理人员及早识别体外循环冠状动脉旁路移植术后机械通气时间延长高危患者并实施预防性护理措施提供借鉴。

Objective To develop a predictive model for assessment of the risk of the patients on prolonged mechanical ventilation after coronary artery bypass grafting with extracorporeal circulation. Methods A convenience sampling method was employed to select 2 334 patients who received the coronary artery bypass grafting（CABG）with extracorporeal circulation in our hospital from January 2021 to December 2023 as the study subjects. Preoperative, intraoperative and postoperative data were collected through structured queries from the electronic medical record system of hospital. The study subjects were randomly divided into a training set （n=1 633）and a validation set（n=701）following a 3：1 ratio. A risk prediction model was established using Logistic regression based on the training set data. Model fit was assessed using Hosmer-Lemeshow test, and predictive performance of the model was evaluated with the area under curve（AUC）of the receiver operating characteristic（ROC）curve. Results A total of 2, 334 patients were included, of whom 215（9.2%）experienced the prolonged mechanical ventilation（>24 hours）. The model developed from the training set identified seven factors that contributed to a prolonged mechanical ventilation：age（OR=1.03）, body mass index（BMI, OR=1.14）, time of extracorporeal circulation（OR=1.01）, intraoperative blood transfusion（OR=4.15）, postoperative serum total bilirubin（OR=1.08）, postoperative serum albumin（OR=0.92）and postoperative re-sternotomy（OR=5.49）. The AUC of the model for prediction of prolonged mechanical ventilation after CABG with extracorporeal circulation was 0.761, with a 95%CI of 0.716-0.806, a maximum Youden index of 0.105, a sensitivity of 77.94%, and a specificity of 64.38%. Validation using the validation set data yielded an AUC of 0.733, with a 95%CI of 0.662-0.804, a sensitivity of 75.32%, a specificity of 57.97%, and a predictive accuracy of 73.61%. Conclusion The risk prediction model developed in this study for prolonged mechanical ventilation after a CABG with extracorporeal circulation demonstrates a good predictive performance. It provides a reference for the nurses to identify the patient in high-risk of prolonged mechanical ventilation after a CABG with extracorporeal circulation and to implement preventive nursing measures.