室间隔缺损(Ventricular Septal Defect, VSD)是临床上最常见的先天性心脏疾病(Congenital heart disease, CHD),不仅可单独发生,也可与其他复杂心脏畸形共存。VSD手术治疗在降低患儿的病死率和提高其生活质量方面具有不可忽视的作用。手术前心力衰竭(Heart Failure, HF)的准确诊断及干预是提高手术成功率、降低术后不良事件发生的关键。在成人心力衰竭诊断中常将脑钠肽(Brain Natriuretic Peptide, BNP)等生物标志物作为诊断和治疗的依据,然而小儿心力衰竭及先天性心脏病,没有任何临床生物标志物作为诊断或治疗的标准指南。在信息时代,基于机器学习(Machine Learning, ML)算法建立的模型可提高对相关危险因素预测的准确性。本文结合相关文献对室间隔缺损术前心力衰竭及术后不良事件发生的预测因素进行总结。Ventricular Septal Defect (VSD) is the most common congenital heart disease (CHD) clinically, which can occur either alone or in combination with other complex heart malformations. Surgical treatment of VSD plays a significant role in reducing mortality and improving the quality of life of affected children. Accurate diagnosis and intervention for heart failure (HF) before surgery are crucial for enhancing surgical success rates and minimizing postoperative adverse events. In adult heart failure diagnosis, biomarkers such as brain natriuretic peptide (BNP) are often used as a basis for diagnosis and treatment. Nevertheless, for pediatric heart failure and congenital heart disease, there are no clinical biomarkers serving as standard guidelines for diagnosis or treatment. In the information era, models based on machine learning (ML) algorithms can improve the accuracy of predicting relevant risk factors. This article summarizes the predictive factors for preoperative heart failure and postoperative adverse events in patients with ventricular septal defects, drawing on relevant literature.
目的 探索食管癌患者术后呼吸衰竭（respiratory failure,RF）的危险因素，构建基于最小绝对收缩和选择算子（least absolute shrinkage and selection operator,LASSO）-logistic回归的预测模型，并对所建模型进行可视化处理。方法 回...
详细信息
目的 探索食管癌患者术后呼吸衰竭（respiratory failure,RF）的危险因素，构建基于最小绝对收缩和选择算子（least absolute shrinkage and selection operator,LASSO）-logistic回归的预测模型，并对所建模型进行可视化处理。方法 回顾性纳入2020—2023年在中山大学附属肿瘤医院甘肃医院胸外科接受手术治疗的食管癌患者，根据术后是否发生RF，将患者分为RF组和非RF（non-respiratory failure,NRF）组。收集两组患者的临床资料，应用LASSO-logistic回归优化模型的特征选择，构建预测模型。基于Bootstrap法重复抽样1 000次对模型进行内部验证。结果 共纳入217例患者，其中RF组24例，男22例、女2例，平均年龄（63.33±9.10）岁；NRF组193例，男161例、女32例，平均年龄（62.14±8.44）岁。LASSO-logistic回归分析显示，一秒率（forced expiratory volume in one second/forced vital capacity,FEV1/FVC）占预计值的百分比（percentage of FEV1/FVC to predicated value,FEV1/FVC%pred）[OR=0.944,95%CI（0.897,0.993）,P=0.026]、术后吻合口瘘[OR=4.106,95%CI（1.457,11.575）,P=0.008]、术后肺部感染[OR=3.776,95%CI（1.373,10.388）,P=0.010]是食管癌术后RF的危险因素。根据上述危险因素构建预测模型，受试者工作特征曲线下面积为0.819[95%CI（0.737,0.901）]。校准曲线Hosmer-Lemeshow检验表明该模型具有良好的拟合优度（P=0.527）。决策曲线显示阈值概率在5%～50%时模型具有良好的临床净收益。结论 FEV1/FVC%pred、术后吻合口瘘、术后肺部感染是食管癌患者术后RF的危险因素，基于LASSO-logistic回归分析法构建的预测模型有望帮助医务人员筛选出高危患者，以便进行早期个体化干预。
同方期刊数据库
维普期刊数据库
暂无评论