脑胶质瘤术后谵妄护理预警模型的构建与验证  

Construction and validation of a nursing early warning model for postoperative delirium in patients with glioma

作　　者：张春丰 褚庆森 石晓梅 杨帆 李迎 Zhang Chunfeng;Chu Qingsen;Shi Xiaomei;Yang Fan;Li Ying(Department of Anesthesia and Surgery,Provincial Hospital Affiliated to Shandong First Medical University,Jinan 250021,China;Department of Neurosurgery,Provincial Hospital Affiliated to Shandong First Medical University,Jinan 250021,China;Department of Respiratory Medicine,Shandong Third Provincial Hospital,Jinan 250031,China)

机构地区：[1]山东第一医科大学附属省立医院麻醉手术科,济南250021 [2]山东第一医科大学附属省立医院神经外科,济南250021 [3]山东省立第三医院呼吸内科,济南250031

出　　处：《中华行为医学与脑科学杂志》2022年第11期996-1001,共6页Chinese Journal of Behavioral Medicine and Brain Science

基　　金：山东省自然科学基金(ZR2020QH108);济南市科技计划(202019088);山东省立医院护理科研项目(HL2021-12)。

摘　　要：目的通过构建列线图护理预测模型分析脑胶质瘤患者术后谵妄发生的相关危险因素并加以验证。方法选取2021年1月至2022年1月山东省立医院神经外科收治的304例脑胶质瘤手术患者作为研究对象,采用留出法分为训练集(234例)和验证集(70例)。以意识模糊评估表(confusion assessment method,CAM)将训练集和验证集患者分为谵妄组和非谵妄组,分析训练集和验证集两组患者的临床资料,并通过Logistic回归分析明确术后谵妄的危险因素,建立脑胶质瘤术后谵妄预警模型。利用校准曲线和ROC曲线分别在训练集及验证集中进行内、外部验证,评价预测模型精确度和区分度。结果多因素Logistic回归分析表明年龄>60岁(OR=4.089,95%CI=1.898~9.103,P<0.001)、糖尿病(OR=2.825,95%CI=1.316~6.186,P=0.008)、高血压病(OR=2.176,95%CI=1.041~4.587,P=0.008)、吸烟(OR=2.432,95%CI=1.063~5.648,P=0.036)、既往癫痫发作史(OR=4.457,95%CI=1.924~10.689,P=0.001)、肺功能差(OR=2.452,95%CI=1.132~5.374,P=0.023)、疼痛视觉模拟量表(visual analog scale,VAS)>7分(OR=3.394,95%CI=1.591~7.456,P=0.002)焦虑或抑郁(OR=2.746,95%CI=1.285~5.976,P=0.010)和手术时长>4 h(OR=2.731,95%CI=1.255~6.062,P=0.012)是脑胶质瘤术后谵妄的独立危险因素,基于上述危险因素建立列线图护理预警模型,内部验证ROC曲线下面积AUC=0.852,外部验证ROC曲线下面积AUC=0.914。结论脑胶质瘤术后谵妄护理预警模型能有效预测脑胶质瘤术后谵妄的发生风险,有一定的临床推广价值。Objective To analyze the risk factors for postoperative delirium in patients with glioma,and construct and validate a nomogram prediction model.Methods A total of 304 glioma surgery patients admitted to the Department of Neurosurgery of Shandong Provincial Hospital from January 2021 to January 2022 were involved in this research.The training set(234 patients)and the validation set(70 patients)were divided according to the leave-out method.Patients in the training set and validation set were divided into delirium and non-delirium groups using the confusion assessment method(CAM)as the criteria.Clinical data of patients in the two study groups in the training set were analyzed.The risk factors for postoperative delirium were clarified by Logistic regression analysis,and an early warning model for postoperative delirium in glioma was established.The calibration curve and ROC curve were used for internal and external validation in the training set and validation set to evaluate the accuracy and discrimination of the prediction model.Results Multivariate Logistic regression analysis showed that age>60(OR=4.089,95%CI=1.898-9.103,P<0.001),diabetes(OR=2.825,95%CI=1.316-6.186,P=0.008),hypertension(OR=2.176,95%CI=1.041-4.587,P=0.008),smoking(OR=2.432,95%CI=1.063-5.648,P=0.036),history of epileptic seizures(OR=4.457,95%CI=1.924-10.689,P=0.001),poor lung function(OR=2.452,95%CI=1.132-5.374,P=0.023),visual analog scale(VAS)>7 points(OR=3.394,95%CI=1.591-7.456,P=0.002),anxiety or depression(OR=2.746,95%CI=1.285-5.976,P=0.010)and operation duration>4 h(OR=2.731,95%CI=1.255-6.062,P=0.012)were the independent risk factors of brain glial postoperative delirium.Based on the above risk factors a nomogram nursing early warning model was established.The area under the ROC curve(AUC)of internal verification was 0.852,and AUC of external verification was 0.914.Conclusion The early-warning model for postoperative delirium in glioma patients can effectively predict the risk of delirium after glioma surgery,and it has certain clinical prom

关 键 词：脑胶质瘤 术后谵妄 危险因素 预测模型 

分 类 号：R473.73[医药卫生—护理学]