Department of Anesthesiology, The Third Affiliated Hospital of Sun Yat-Sen University, No. 600 Tianhe Road, Guangzhou, 10630, Guangdong, People's Republic of China.
Guangzhou AID Cloud Technology Co., LTD, Guangzhou, People's Republic of China.
Respir Res. 2021 Mar 31;22(1):94. doi: 10.1186/s12931-021-01690-3.
Pneumonia is the most frequently encountered postoperative pulmonary complications (PPC) after orthotopic liver transplantation (OLT), which cause high morbidity and mortality rates. We aimed to develop a model to predict postoperative pneumonia in OLT patients using machine learning (ML) methods.
Data of 786 adult patients underwent OLT at the Third Affiliated Hospital of Sun Yat-sen University from January 2015 to September 2019 was retrospectively extracted from electronic medical records and randomly subdivided into a training set and a testing set. With the training set, six ML models including logistic regression (LR), support vector machine (SVM), random forest (RF), adaptive boosting (AdaBoost), extreme gradient boosting (XGBoost) and gradient boosting machine (GBM) were developed. These models were assessed by the area under curve (AUC) of receiver operating characteristic on the testing set. The related risk factors and outcomes of pneumonia were also probed based on the chosen model.
591 OLT patients were eventually included and 253 (42.81%) were diagnosed with postoperative pneumonia, which was associated with increased postoperative hospitalization and mortality (P < 0.05). Among the six ML models, XGBoost model performed best. The AUC of XGBoost model on the testing set was 0.734 (sensitivity: 52.6%; specificity: 77.5%). Pneumonia was notably associated with 14 items features: INR, HCT, PLT, ALB, ALT, FIB, WBC, PT, serum Na, TBIL, anesthesia time, preoperative length of stay, total fluid transfusion and operation time.
Our study firstly demonstrated that the XGBoost model with 14 common variables might predict postoperative pneumonia in OLT patients.
肺炎是肝移植（OLT）后最常见的术后肺部并发症（PPC），导致高发病率和死亡率。我们旨在使用机器学习（ML）方法为 OLT 患者开发一种预测术后肺炎的模型。
从电子病历中回顾性提取 2015 年 1 月至 2019 年 9 月在中山大学附属第三医院接受 OLT 的 786 名成年患者的数据，并将其随机分为训练集和测试集。使用训练集，开发了包括逻辑回归（LR）、支持向量机（SVM）、随机森林（RF）、自适应增强（AdaBoost）、极端梯度增强（XGBoost）和梯度提升机（GBM）在内的 6 个 ML 模型。使用测试集上的接收者操作特征曲线（AUC）评估这些模型。还根据选定的模型探究了肺炎的相关风险因素和结果。
最终纳入 591 名 OLT 患者，其中 253 名（42.81%）被诊断为术后肺炎，与术后住院时间延长和死亡率增加有关（P < 0.05）。在六个 ML 模型中，XGBoost 模型表现最佳。XGBoost 模型在测试集上的 AUC 为 0.734（灵敏度：52.6%；特异性：77.5%）。肺炎与 14 项特征明显相关：INR、HCT、PLT、ALB、ALT、FIB、WBC、PT、血清 Na、TBIL、麻醉时间、术前住院时间、总液体输注量和手术时间。
本研究首次表明，XGBoost 模型结合 14 个常见变量可能预测 OLT 患者术后肺炎。
