2022 年第 10 期 第 17 卷

基于血清25-羟基维生素D水平构建老年患者硬质支气管镜术后谵妄预测模型及验证

Construction and validation of predictive model for delirium after rigid bronchoscopy in elderly patients based on serum 25-hydroxy vitamin D level

作者：黄媛媛张维任心怡何杰

英文作者：Huang Yuanyuan Zhang Wei Ren Xinyi He Jie

单位：成都医学院第一附属医院呼吸与危重症医学科，成都610500

英文单位：Department of Respiratory and Critical Care Medicine the First Affiliated Hospital of Chengdu Medical College Chengdu 610500 China

关键词：谵妄；硬质支气管镜；麻醉；25-羟基维生素D

英文关键词：Delirium;Rigidbronchoscopy;Anesthesia;25-hydroxyvitaminD

• 摘要：

• 目的 分析老年患者硬质支气管镜术后谵妄的危险因素，基于血清25-羟基维生素D［25-(OH)D］水平构建患者术后谵妄预测模型并进行验证。方法 选取2015年7月至2020年7月在成都医学院第一附属医院呼吸与危重症医学科接受全身麻醉下硬质支气管镜手术的老年患者1 052例。通过随机数字表法按照7∶3的比例分为建模组（736例）和验证组（316例）。根据术后1～3 d是否发生谵妄分为术后谵妄组和非术后谵妄组。采用Logistic回归方法分析术后谵妄的危险因素。在建模人群中，基于多因素Logistic回归分析筛选的危险因素，建立预测患者术后谵妄的列线图模型。分别在建模人群和验证人群中对预测模型进行区分度和校准度的验证。结果 年龄（比值比=1.064,95%置信区间：1.012～1.146，P=0.024）、血清25-(OH)D（比值比=2.596,95%置信区间：1.124～6.985，P=0.025）、术前简易智力状态检查量表（MMSE）评分（比值比=0.913,95%置信区间：0.817～0.977，P=0.004）、麻醉时间（比值比=4.884,95%置信区间：1.308～11.103，P=0.031）、血清白蛋白（比值比=7.262,95%置信区间：2.865～18.991，P=0.024）、气道恶性肿瘤（比值比=1.847，95%置信区间：1.218～3.023，P=0.001）是老年患者硬质支气管镜术后发生谵妄的独立危险因素。纳入上述6个独立危险因素建立老年患者硬质支气管镜术后谵妄发生风险的列线图预测模型，建模人群验证模型的受试者工作特征曲线下面积为0.761（95%置信区间：0.689～0.822），C指数为0.788，验证人群曲线下面积为0.749（95%置信区间：0.653～0.815），C指数为0.765。校正曲线显示，在建模人群和验证人群中，该列线图模型的预测结果与实际结果之间具备良好的一致性。结论 高龄、维生素D缺乏，气道恶性肿瘤，术前低MMSE评分、低血清白蛋白和长时间麻醉是老年患者硬质支气管镜术后发生谵妄的独立危险因素。基于血清25-(OH)D建立列线图预测模型有助于临床早期筛选并识别术后发生谵妄的高危患者。

• Objective  To analyze the risk factors of delirium after rigid bronchoscopy in elderly patients, to construct a predictive model of delirium after rigid bronchoscopy in elderly patients based on the level of serum 25-hydroxy vitamin D ［25-(OH)D］ and to verify it. Methods A total of 1 052 elderly patients who underwent rigid bronchoscopy under general anesthesia in Department of Respiratory and Critical Care Medicine, the First Affiliated Hospital of Chengdu Medical College from July 2015 to July 2020 were enrolled in this study. Patients were divided into modeling group (736 cases) and validation group (316 cases) by random number table method according to the ratio of 7∶3. Patients were divided into postoperative delirium group and non-postoperative delirium group according to whether delirium occurred 1-3 d after surgery. Logistic regression analysis was used to analyze the risk factors for postoperative delirium. In the modeling population, a nomogram model for predicting postoperative delirium in patients was developed based on the risk factors screened by multivariate Logistic regression analysis. The prediction model were validated for discrimination and calibration in the modeling population and validation population, respectively. Results The age［odds ratio(OR)=1.064,95% confidence interval(CI)：1.012-1.146，P=0.024］, serum 25-(OH)D（OR=2.596,95% CI：1.124-6.985，P=0.025）, preoperative mini mental status examination scale (MMSE) score（OR=0.913,95% CI：0.817-0.977，P=0.004）, duration of anesthesia（OR=4.884,95% CI：1.308-11.103，P=0.031）, serum albumin（OR=7.262,95% CI：2.865-18.991，P=0.024）, and airway malignancy（OR=1.847，95% CI：1.218-3.023，P=0.001） were independent risk factors for the occurrence of delirium after rigid bronchoscopy in elderly patients.  A nomogram prediction model for the risk of delirium after rigid bronchoscopy in elderly patients was established after incorporating the above six independent risk factors. The area under the receiver operating characteristic curve of the modeling population for verifying this model was 0.761 (95% CI: 0.689-0.822) with C-index of 0.788, and the area under the receiver operating characteristic curve for the validation population was 0.749 (95% CI: 0.653-0.815) with C-index of 0.765. The calibration curve showed high concordance between the predicted and actual Results of this nomogram model in both the modeling and validation populations. Conclusions Advanced age, vitamin D deficiency, airway malignancy, low preoperative MMSE score, low serum albumin and prolonged anesthesia are independent risk factors for delirium after rigid bronchoscopy in elderly patients. Establishing a nomograph model based on serum 25-(OH)D can help clinical early screening and identification of elderly patients at high risk for postoperative delirium.