成人术后发生寒战的影响因素分析及预测模型构建  

Analysis of influencing factors and construction of prediction model for postoperative shivering in adults

作　　者：马世发[1] 刘鑫 翟瑞金 陆春秀[3] 范科慧[3] MA Shifa;LIU Xin;ZHAI Ruijin;LU Chunxiu;FAN Kehui(Central Sterile Supply Department,The First People's Hospital of Nanning,Nanning 530022,China;不详)

机构地区：[1]南宁市第一人民医院消毒供应中心,广西南宁530022 [2]南宁市第一人民医院护理部,广西南宁530022 [3]南宁市第一人民医院手术室,广西南宁530022

出　　处：《山东医药》2025年第3期116-121,共6页Shandong Medical Journal

基　　金：广西壮族自治区卫生健康委员会科研课题(Z20201277)。

摘　　要：目的分析成人术后发生寒战的影响因素,构建成人术后发生寒战的预测模型。方法接受手术治疗的患者613例,按8∶2的比例随机分成建模组490例、验证组123例。统计建模组患者术后寒战发生情况,并收集患者术后发生寒战的有关资料,包括性别、年龄、术中液体冲洗量、保暖措施、术后疼痛分级、手术分级、手术类型、手术持续时间、麻醉类型、术中液体入量、术中出血量,以建模组中单因素分析有统计学差异的指标为自变量,以成人术后发生寒战为因变量,构建Logistic回归模型,分析成人术后发生寒战的影响因素。采用R4.3.0软件和rms方程包构建成人术后发生寒战的预测模型,采用Bootstrap方法对模型进行内部验证和验证集验证。建模组和验证组采用受试者工作特征(ROC)曲线、校正曲线及临床决策曲线验证模型的区分度、一致性和临床有效性。结果年龄、手术持续时间、术中液体入量、术中出血量、术后疼痛分级、手术分级、保暖措施是成人术后发生寒战的独立影响因素(P均<0.05),据此构建了建模组成人术后发生寒战的预测模型。建模组预测模型的AUC值为0.760(95%CI:0.690~0.830),验证组为0.702(95%CI:0.556~0.848),均提示该预测模型的区分度良好。建模组和验证组成人术后发生寒战预测模型的校准曲线中,预测模型的预测发生率和实际发生率之间有很好的一致性。建模组和验证组成人术后发生寒战预测模型的临床决策曲线中,在一定的阈值概率范围内,基于该模型预测结果进行早期临床监测和干预,相较于对所有人干预或不对任何人干预,能获得更高的净收益,提示该模型在建模组和验证组中均具有一定的临床有效性。结论年龄、手术持续时间、术中液体入量、术中出血量、术后疼痛分级、手术分级、保暖措施是成人术后发生寒战的独立影响因素。成功构建了建模组Objective To analyze the influencing factors of postoperative shivering in adults and to construct a prediction model for postoperative shivering in adults.Methods A total of 613 patients who underwent surgical treatment were randomly divided into the model group of 490 cases and validation group of 123 cases at a ratio of 8:2.The occurrence of postoperative shivering in the model group was counted.Data related to postoperative shivering,including gender,age,volume of intraoperative fluid irrigation,warming measures,postoperative pain grading,surgical grading,type of surgery,duration of surgery,type of anesthesia,intraoperative fluid intake,and intraoperative blood loss,were collected.Taking the indicators with statistically significant differences in the univariate analysis of the model group as the independent variables and the occurrence of postoperative shivering in adults as the dependent variable,we constructed a Logistic regression model to analyze the influencing factors of postoperative shivering in adults.The R4.3.0 software and the rms package were used to construct a prediction model for postoperative shivering in adults.The Bootstrap method was applied for internal validation of the model and validation set validation.The receiver operating characteristic(ROC)curve,calibration curve,and clinical decision curve were used in the model group and the validation group to verify the discrimination,consistency,and clinical effectiveness of the model.Results Age,duration of surgery,intraoperative fluid intake,intraoperative blood loss,postoperative pain grading,surgical grading,and warming measures were independent influencing factors for postoperative shivering in adults(all P<0.05).Based on this,a prediction model for postoperative shivering in adults in the model group was constructed.The prediction model demonstrated good discrimination,with AUC values of 0.760(95%CI:0.690-0.830)in the model group and 0.702(95%CI:0.556-0.848)in the validation group.In the calibration curves of the prediction model for po

关 键 词：术后并发症 术后寒战 成人 影响因素 预测模型 

分 类 号：R734.2[医药卫生—肿瘤]