摘要
目的探讨重度颅脑损伤(STBI)患者术后认知障碍(POCD)关联因素,构建回归预测模型。方法选取南京鼓楼医院神经外科2021年8月至2023年8月STBI手术患者162例,依据术后15 d内POCD发生情况分为POCD组、对照组。比较两组临床资料,采用多因素Logistic回归模型确定POCD发生独立影响因素,依据多因素分析的回归系数构建回归预测模型,通过H-L检验判定预测模型校准能力,受试者操作特征曲线(ROC)评价预测模型的预测效能。结果162例STBI患者中发生POCD52例(32.10%),归为POCD组;未发生POCD 110例(67.90%),归为对照组。经单因素分析显示,两组年龄、体质量指数、文化程度、颅脑损伤部位、损伤部位数量、高血压、血清S100β、神经元特异性烯醇化酶(NSE)水平比较,差异有统计学意义(P<0.05);两组性别、颅脑损伤原因、受伤至入院治疗时间、入院时格拉斯哥昏迷量表(GCS)评分、糖尿病、手术方式比较,差异无统计学意义(P>0.05)。多因素Logistic回归分析发现,年龄、血清S100β、NSE水平均为STBI患者POCD的独立危险因素(OR分别为1.156、17.350、1.178),文化程度为独立保护因素(OR为0.405)(P<0.05)。依据独立影响因素回归系数构建Logistic回归预测模型,Logit(P)=-15.619+0.145×年龄(岁)-0.903×文化程度(大专及以上=3,高中/中专=2,初中=1,小学=0)+2.854×血清S100β(μg·L^(-1))+0.164×血清NSE(μg·L^(-1))。H-L检验显示,χ^(2)=13.959,P=0.083,模型校准能力较好,构建有效。绘制ROC曲线显示,模型预测STBI患者POCD(AUC=0.872,95%CI:0.810~0.933),预测敏感度为75.00%,特异度为91.82%。结论STBI患者POCD关联因素包括年龄、文化程度、血清S100β、NSE,依据上述因素构建的回归预测模型预测价值良好,可为POCD防范提供依据。
Objective To explore the related factors of postoperative cognitive dysfunction(POCD)in patients with severe traumatic brain injury(STBI),and to construct a regression prediction model.Methods 162 patients with STBI surgery in our hospital from August 2021 to August 2023 were selected and divided into POCD group and control group according to the occurrence of POCD within 15 days after operation.The clinical data of the two groups were compared.Multivariate Logistic regression model was used to determine the independent influencing factors of POCD.The regression prediction model was constructed based on the regression coefficient of multivariate analysis.The goodness of fit of the prediction model was determined by H-L test,and the prediction efficiency of the prediction model was evaluated by receiver operating characteristic curve(ROC).Results Among 162 patients with STBI,52(32.10%)had POCD,which was classified as POCD group.There were 110 cases(67.90%)without POCD,which were classified as the control group.Univariate analysis showed that there were statistically significant differences in age,body mass index,education level,traumatic brain injury site,number of injury sites,hypertension,serum S100βand neuron-specific enolase(NSE)levels between the two groups(P<0.05).There were no significant differences in gender,cause of traumatic brain injury,time from injury to admission,Glasgow Coma Scale(GCS)score at admission,diabetes mellitus and surgical methods between the two groups(P>0.05).Multivariate Logistic regression analysis showed that age,serum S100βand NSE levels were independent risk factors for POCD in STBI patients(OR=1.156,17.350 and 1.178,respectively),and education level was an independent protective factor(OR=0.405)(P<0.05).Logistic regression prediction model was constructed according to the regression coefficient of independent influencing factors,Logit(P)=-15.619+0.145×age(years old)-0.903×educational level(junior college and above=3,senior high school/technical secondary school=2,junior high school=1,primary school=0)+2.854×serum S100β(μg·L^(-1))+0.164×serum NSE(μg·L^(-1)).The H-L test showed thatχ^(2)=13.959,P=0.083,the goodness of fit of the model was good and the construction was effective.ROC curve showed that the model for predicting POCD in STBI patients(AUC=0.872,95%CI:0.810-0.933),the predictive sensitivity was 75.00%,and the specificity was 91.82%.Conclusion The related factors of POCD in STBI patients include age,education level,serum S100βand NSE.The regression prediction model based on the above factors has good predictive value and can provide a basis for POCD prevention.
作者
杭晓慧
季翠玲
杭代
Hang Xiaohui;Ji Cuiling;Hang Dai(Department of Neurology,the Nanjing Gulou Hospital,Jiangsu 210000,China)
出处
《脑与神经疾病杂志》
2025年第7期452-456,共5页
Journal of Brain and Nervous Diseases
基金
南京鼓楼医院临床研究及横向科研项目(2023-B1420)。
关键词
重度颅脑损伤
术后认知障碍
预测模型
Severe traumatic brain injury
Postoperative cognitive dysfunction
Prediction model
