|
| 基于随机森林算法建立老年患者脊髓麻醉后低血压的预测模型 |
| Prediction model for spinal anesthesia-induced hypotension in elderly patients based on random forest algorithm |
| |
| DOI：10.12089/jca.2025.11.001 |
| 中文关键词: 随机森林 老年 脊髓麻醉 低血压 超声心动图 |
| 英文关键词: Random forest Aged Spinal anesthesia Hypotension Echocardiography |
| 基金项目: |
|
| 摘要点击次数: 1320 |
| 全文下载次数: 351 |
| 中文摘要: |
目的：筛选老年患者脊髓麻醉(腰麻)后低血压（SAIH）的独立危险因素，并基于随机森林算法建立老年患者SAIH预测模型。
方法：回顾性收集2023年1月至2024年9月行择期腰麻手术的778例老年患者的临床资料，包括一般资料、术前超声心动图参数和术中情况。将患者按照8∶2的比例随机分为训练集（n=622）和验证集（n=156），通过LASSO回归、Logistic回归分析筛选老年患者SAIH的影响因素，采用随机森林算法建立并验证SAIH预测模型。采用受试者工作特征（ROC）曲线下面积（AUC）评价随机森林模型的预测性能。
结果：有322例（41.4%）患者发生SAIH，其中训练集256例（41.2%），验证集66例（42.3%）。年龄、BMI、麻醉平面、室间隔舒张末期厚度（IVSd）、二尖瓣环舒张早期运动速度（e′）、二尖瓣舒张早期血流速度与二尖瓣环舒张早期运动速度的比值（E/e′）是老年患者SAIH的影响因素。包含e′、E/e′、麻醉平面的随机森林模型在预测性能上表现最佳，模型在训练集中的AUC为0.940（95%CI 0.927～0.952），敏感性0.712，特异性0.922，准确性0.838；在验证集中的AUC为0.888（95%CI 0.829～0.934），敏感性0.611，特异性0.881，准确性0.756。
结论：老年患者SAIH的影响因素包括年龄、BMI、麻醉平面、IVSd、e′、E/e′，基于随机森林算法建立的SAIH模型具有较好的预测性能。 |
| 英文摘要: |
Objective: To identify independent risk factors for spinal anesthesia (SA)-induced hypotension (SAIH) in elderly patients and establish a predictive model based on the random forest algorithm.
Methods: A retrospective analysis was conducted on 778 elderly patients undergoing elective surgery under spinal anesthesia from January 2023 to September 2024. Clinical data included demographic characteristics, preoperative echocardiographic parameters, and intraoperative date. Patients were randomly assigned to the training set (n = 622) and validation set (n = 156) at a ratio of 8∶2. Influencing factors for SAIH in elderly patients were screened using LASSO regression and logistic regression, followed by the development and validation of a random forest-based predictive model. Model performance was evaluated by the area under the receiver operating characteristic (ROC) curve.
Results: A total of 322 patients (41.4%) developed SAIH, including 256 patients (41.2%) in the training set and 66 patients (42.3%) in the validation set. Influencing factors for SAIH in elderly patients included age, BMI, spinal anesthesia level, interventricular septal diastolic thickness (IVSd), early diastolic mitral annular velocity (e′), and ratio of early diastolic mitral flow velocity to early diastolic mitral annular velocity (E/e′). The random forest model incorporating e′, E/e′, and spinal anesthesia level demonstrated the best predictive performance. In the training set, the model had an area under the cure (AUC) of 0.940 (95% CI 0.927-0.952), with a sensitivity of 0.712, specificity of 0.922, and accuracy of 0.838. In the validation set, the AUC was 0.888 (95% CI 0.829-0.934), with a sensitivity of 0.611, specificity of 0.881, and accuracy of 0.756.
Conclusion: Influencing factors for SAIH in elderly patients include age, BMI, sensory block level, IVSd, e′, and E/e′, and the random forest-based predictive model exhibits robust performance. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
