2023 年第 2 期 第 18 卷

体外循环术后患者发生急性呼吸窘迫综合征危险因素分析及列线图预测模型构建

Risk factors analysis of acute respiratory distress syndrome after cardiopulmonary bypass and construction of nomogram prediction model

作者：刘妍石秀梅朱光发

英文作者：Liu Yan Shi Xiumei Zhu Guangfa

单位：首都医科大学附属北京安贞医院感染科北京市心肺血管疾病研究所，北京100029

英文单位：Department of Infectious Disease Beijing Anzhen Hospital Capital Medical University Beijing Institute of Heart Lung and Blood Vessel Diseases Beijing 100029 China

关键词：体外循环术；急性呼吸窘迫综合征；预测模型；列线图

英文关键词：Cardiopulmonarybypass;Acuterespiratorydistresssyndrome;Predictionmodel；Nomogram

• 摘要：

• 目的 探讨体外循环术后患者发生急性呼吸窘迫综合征（ARDS）的危险因素并构建列线图预测模型。方法 本研究为回顾性病例对照研究。纳入2005年1月至2015年12月入住首都医科大学附属北京安贞医院心脏外科在体外循环辅助下行心脏或大血管手术，并于术后发生ARDS的患者145例为ARDS组。根据每例ARDS患者的手术年份及年龄，随机匹配2例体外循环辅助下行心脏或大血管手术后未发生ARDS的患者作为对照组（290例）。应用Logistic回归方法分析体外循环术后ARDS的危险因素，并应用列线图方法构建体外循环术后ARDS预测模型。结果 多因素Logistic回归分析结果显示，低术前血红蛋白水平（比值比=0.98，95%置信区间：0.96～0.99，P=0.001）、手术时间增加（比值比=1.26，95%置信区间：1.05～1.51，P=0.011）、术后第1天高急性生理学与慢性健康状况评分系统Ⅱ（APACHEⅡ）评分（比值比=1.14，95%置信区间：1.08～1.23，P＜0.001）、吸烟（比值比=2.65，95%置信区间：1.14～6.05，P=0.023）、瓣膜联合冠状动脉旁路移植术（CABG）手术（比值比=2.59，95%置信区间：1.11～6.01,P=0.027）为体外循环辅助下行心脏或大血管手术患者术后发生ARDS的独立危险因素。将上述因素全部纳入预测模型构建列线图，该预测模型的曲线下面积为0.834，最佳阈值为0.334时，特异度为0.814，敏感度为0.748。结论 低术前血红蛋白水平、手术时间增加、高APACHEⅡ评分、吸烟、瓣膜联合CABG手术为体外循环术后发生ARDS的独立危险因素。以这些危险因素构建的列线图预测模型可用于体外循环术后患者ARDS的早期识别。

• Objective To investigate the risk factors of acute respiratory distress syndrome (ARDS) after cardiopulmonary bypass, and to construct a nomogram prediction model. Methods  This was a retrospective case-control study. Patients admitted to Department of Cardiac Surgery, Beijing Anzhen Hospital, Capital Medical University from January 2005 to December 2015 were enrolled. Patients underwent cardiac or macrovascular surgery assisted by cardiopulmonary bypass, and 145 patients who developed ARDS postoperatively were divided into the ARDS group. According to the year of surgery and age of each patient with ARDS, 2 matched patients who did not develop ARDS after cardiac or macrovascular surgery assisted by cardiopulmonary bypass were randomly divided into control group (290 cases). Logistic regression analysis was used to analyze the risk factors for ARDS after cardiopulmonary bypass, and nomogram  was applied to construct a prediction model for ARDS after cardiopulmonary bypass. Results  Multivariate Logistic regression analysis showed that low preoperative hemoglobin level (odds ratio=0.98, 95% confidence interval: 0.96-0.99, P=0.001), increased operation time (odds ratio=1.26, 95% confidence interval: 1.05-1.51, P=0.011), high acute physiology and chronic health evaluation scoring system Ⅱ (APACHEⅡ) score (odds ratio=1.14, 95% confidence interval: 1.08-1.23, P＜0.001), smoking (odds ratio=2.65, 95% confidence interval: 1.14-6.05, P=0.023), and valve surgery in combination with coronary artery bypass grafting (CABG) (odds ratio=2.59, 95% confidence interval: 1.11-6.01, P=0.027) were independent risk factors of postoperative ARDS in patients undergoing cardiac or macrovascular surgery assisted by cardiopulmonary bypass. The nomogram was constructed by incorporating all of the above factors into the prediction model, and the area under the curve of this prediction model was 0.834, with a specificity of 0.814 and a sensitivity of 0.748 at an optimal threshold of 0.334. Conclusions  Low preoperative hemoglobin level, increased operation time, high APACHEⅡ score, smoking and valve surgery in combination with CABG were independent risk factors of ARDS after cardiopulmonary bypass. The nomogram prediction model constructed with these risk factors can be used for the early identification of ARDS in patients after cardiopulmonary bypass.