心脏骤停复苏后早期发生急性肾损伤预测模型构建及验证  

Construction and validation of a predictive model for early acute kidney injury in patients with cardiac arrest after cardiopulmonary resuscitation

作　　者：王晋祥[1] 华罗刚 余慕明[1] 王力军[1] 靳衡[1] 续国武[1] Wang Jinxiang;Hua Luogang;Yu Muming;Wang Lijun;Jin Heng;Xu Guowu(Department of Emergency Medicine,Tianjin Medical University General Hospital,Tianjin 300052,China)

机构地区：[1]天津医科大学总医院急诊医学科,天津300052

出　　处：《中华急诊医学杂志》2025年第1期17-24,共8页Chinese Journal of Emergency Medicine

基　　金：天津市教委科研计划项目(2021KJ211)。

摘　　要：目的构建预测心脏骤停(cardiac arrest,CA)复苏后患者早期发生急性肾损伤(acute kidney injury,AKI)的列线图模型,并验证其早期预测的有效性。方法回顾性纳入2016年2月至2023年9月在天津医科大学总医院急诊抢救室收治的CA经心肺复苏(cardiopulmonary resuscitation,CPR)且年龄在18岁以上的患者,收集患者的一般资料、基础疾病、复苏相关指标及首次实验室检测结果等。按照7∶3的比例随机将患者分为训练组和验证组。AKI诊断参照改善全球肾脏预后组织诊断标准。采用单因素和多因素Logistic回归分析CA复苏后患者发生AKI的影响因素并构建列线图模型。通过受试者工作特征曲线的曲线下面积(area under the curve,AUC)评估预测效能,采用校准曲线、决策曲线及临床影响曲线对模型进行评价,同时采用Bootstrap法和交叉验证法进行内部验证。结果共纳入527例CA患者,其中230例发生AKI,AKI发生率为43.6%。训练组与验证组临床基线资料差异无统计学意义(均P>0.05),说明两组数据具有可比性。多因素Logistic分析发现年龄(OR=1.346,95%CI:1.197~1.543,P<0.001)、CA至CPR时间(OR=2.214,95%CI:1.512~3.409,P=0.016)、肾上腺素用量(OR=1.921,95%CI:1.383~2.783,P=0.004)、APACHE-Ⅱ评分(OR=1.531,95%CI:1.316~1.820,P<0.001)、基线肌酐(OR=1.137,95%CI:1.090~1.196,P<0.001)、乳酸(OR=2.558,95%CI:1.680~4.167,P<0.001)是发生AKI的独立危险因素,初始可除颤心律(OR=0.214,95%CI:0.051~0.759,P=0.023)是发生AKI的保护性因素。基于以上变量构建列线图预测模型。训练组AUC为0.943(95%CI:0.921~0.965),验证组AUC为0.917(95%CI:0.874~0.960),该模型具有良好的区分度、校准度、临床有效性和应用价值。结论基于年龄、CA至CPR时间、初始可除颤心律、肾上腺素用量、APACHE-Ⅱ评分、基线肌酐及乳酸构建的列线图对CA复苏后早期发生AKI具有良好的预测价值,可为早期识别AKI与精准干预提供新的策略。Objective To construct a nomogram model for predicting the occurrence of acute kidney injury(AKI)in patients with cardiac arrest(CA)after cardiopulmonary resuscitation(CPR),and to verify its validity for early prediction.Methods The study retrospectively included patients aged 18 years and older who received CPR for CA and were admitted to the emergency room of Tianjin Medical University General Hospital from February 2016 to September 2023.The general information,underlying diseases,resuscitation related indicators,and first laboratory test results of patients were collected.The patients were randomly divided into training and validation groups at a ratio of 7:3.AKI diagnosis was based on the diagnostic criteria of the Kidney Disease Improving Global Outcomes.Univariate and multivariate logistic regression models were used to identify independent risk factors for AKI in patients with cardiac arrest,and a nomogram was constructed on the basis of the independent risk factors.The predictive performance was evaluated by the area under the curve(AUC)of the receiver operating characteristic.The calibration curve,decision curve and clinical impact curve were used to evaluate the model.Bootstrap and cross validation methods were used for internal validation.Results A total of 527 patients with cardiac arrest were included in the study,230 patients developed AKI,with an AKI incidence of 43.6%.There was no statistically significant difference in clinical baseline data between the training and validation groups(all P>0.05),indicating comparability between the two groups of data.Multivariate logistic analysis revealed that age(OR=1.346,95%CI:1.197-1.543,P<0.001),CA to CPR time(OR=2.214,95%CI:1.512-3.409,P=0.016),adrenaline dosage(OR=1.921,95%CI:1.383-2.783,P=0.004),APACHE-Ⅱscore(OR=1.531,95%CI:1.316-1.820,P<0.001),baseline creatinine(OR=1.137,95%CI:1.090-1.196,P<0.001),and lactate(OR=2.558,95%CI:1.680-4.167,P<0.001)were the independent risk factors for AKI in patients with cardiac arrest.Initial defibrillable rhythm(OR=0.

关 键 词：心脏骤停 心肺复苏 急性肾损伤 列线图 预测模型 

分 类 号：R541.78[医药卫生—心血管疾病] R692[医药卫生—内科学]