脑挫裂伤合并脑疝患者开颅术后30 d预后不良的列线图预测模型构建与评估  

Construction and evaluation of nomogram prediction model for poor prognosis 30 days after craniotomy in patients with brain contusion and laceration complicated with brain hernia

作　　者：申海龙[1] 陈云庆[1] 张龙[1] 胡洋洋[1] Shen Hailong;Chen Yunqing;Zhang Long;Hu Yangyang(Department of Neurosurgery, the Second Affiliated Hospital of Hebei North University, Hebei Province,Zhangjiakou 075000,China)

机构地区：[1]河北北方学院附属第二医院神经外科,张家口075000

出　　处：《疑难病杂志》2021年第12期1234-1239,共6页Chinese Journal of Difficult and Complicated Cases

基　　金：河北省医学科学研究课题计划项目(20191871)。

摘　　要：目的分析脑挫裂伤合并脑疝患者开颅术后30 d预后不良的危险因素,构建列线图预测模型,同时评估模型的区分度和一致性。方法选取2016年1月—2021年4月河北北方学院附属第二医院神经外科开颅手术治疗脑挫裂伤合并脑疝患者136例,依据术后30 d改良Rankin量表得分,将患者分成预后良好组(≤2分)62例与预后不良组(>2分)74例。比较2组临床资料;采用Logistic回归分析影响脑挫裂伤合并脑疝患者开颅术后30 d预后不良的危险因素;构建预测脑挫裂伤合并脑疝患者开颅术后30 d预后不良的列线图模型,且使用受试者工作特征曲线(ROC)及校准曲线验证列线图模型的区分度和一致性。结果预后良好组与预后不良组在入院时间、脑疝分期、瞳孔改变、血压异常、患高血压、入院时格拉斯哥昏迷量表(GCS)评分、入院时脑挫裂伤部位、出血量、入住ICU时间、甘露醇应用时间、入院前是否应用过抗凝药物、手术时机方面比较差异有统计学意义(t/χ^(2)/P=32.181/0.000、12.016/0.002、5.028/0.025、4.399/0.036、4.980/0.026、9.263/0.002、6.593/0.010、4.943/0.026、4.594/0.032、9.066/0.003、5.540/0.019、11.083/0.001);Logistic回归分析结果显示,入院时GCS评分<8分、入院时多发脑挫裂伤、入院时出血量≥30 ml、入住ICU时间≥1周、应用甘露醇≥2周、入院前应用过抗凝药物、入院24 h后行手术治疗是影响脑挫裂伤合并脑疝患者开颅术后30 d预后不良的危险因素[OR(95%CI)=2.757(1.152~6.600)、2.902(1.182~7.123)、5.935(2.246~15.683)、3.036(1.206~7.643)、4.164(1.648~10.517)、4.017(1.592~10.139)、4.943(1.821~13.417)];构建的列线图预测模型具有较好的区分度(ROC曲线下面积为0.841,95%CI 0.776~0.906)和一致性(Hosmer-Lemeshow拟合优度检验χ^(2)=7.087,P=0.420)。结论基于影响脑挫裂伤合并脑疝患者开颅术后30 d预后不良的危险因素而构建的列线图预测模型,区分度、准确度良�Objective To analyze the risk factors of poor prognosis 30 days after craniotomy in patients with brain contusion and laceration combined with brain hernia,construct a nomogram prediction model,and evaluate the differentiation and consistency of the model.Methods From January 2016 to April 2021,136 patients with brain contusion and laceration complicated with brain hernia were treated by neurosurgical craniotomy in the Second Affiliated Hospital of Hebei North University.According to the score of modified Rankin scale 30 days after operation,the patients were divided into 62 cases in the good prognosis group(≤2 points)and 74 cases in the poor prognosis group(>2 points).The clinical data of the two groups were compared;Logistic regression analysis was used to analyze the risk factors affecting the poor prognosis 30 days after craniotomy in patients with brain contusion and laceration complicated with brain hernia;A nomogram model was constructed to predict the poor prognosis of patients with brain contusion and laceration combined with brain hernia 30 days after craniotomy,and the discrimination and consistency of the nomogram model were verified by subject operating characteristic curve(ROC)and calibration curve.Results There were significant differences in admission time,cerebral hernia stage,pupil changes,abnormal blood pressure,hypertension,Glasgow Coma Scale(GCS)score,brain contusion and laceration site,bleeding volume,ICU admission time,mannitol application time,anticoagulant before admission and operation time between the good prognosis group and the poor prognosis group(t/χ^(2)/P=32.181/0.000,12.016/0.002,5.028/0.025,4.399/0.036,4.980/0.026,9.263/0.002,6.593/0.010,4.943/0.026,4.594/0.032,9.066/0.003,5.540/0.019,11.083/0.001)。Logistic regression analysis showed that GCS score<8 points at admission,multiple cerebral contusion and laceration at admission,bleeding volume≥30 ml at admission,admission time≥1 week,mannitol≥2 weeks,anticoagulant drugs before admission and surgical treatment 24 hours aft

关 键 词：脑挫裂伤 脑疝 开颅术 预后不良 危险因素 列线图 

分 类 号：R651.15[医药卫生—外科学]