应用ROC曲线评价无创心输出量监测对急性呼吸困难鉴别诊断的临床价值  

Assessing the clinical value of non-invasive cardiac output monitoring for differential diagnosis of acute dyspnea by ROC curve

作　　者：钱何布 郁莉莉 姚峰[1] 邹晗 刘军[2] QIAN Hebu;YU Lili;YAO Feng;ZOU Han;LIU Jun(Department of Critical Care Medicine,Suzhou Ninth Hospital Affiliated to Soochow University,Jiangsu,Suzhou 215200,China;Department of Critical Care Medicine,the Affiliated Suzhou Hospital of Nanjing Medical University,Jiangsu,Suzhou 215001,China)

机构地区：[1]苏州大学附属苏州九院重症医学科,江苏苏州215200 [2]南京医科大学附属苏州医院重症医学科,江苏苏州215001

出　　处：《中国医药科学》2023年第9期165-168,共4页China Medicine And Pharmacy

基　　金：江苏省科技项目(BE2021660)。

摘　　要：目的探讨应用ROC曲线评价无创心输出量监测(NICOM)技术对急性呼吸困难鉴别诊断的临床价值。方法回顾性分析2019年6月至2021年12月,入住苏州大学附属苏州九院重症医学科急性呼吸困难患者80例,根据病因分成心源性呼吸困难组(AHF组,38例)和非心源性呼吸困难组(NAHF组,42例),比较两组患者临床资料、血清B型脑钠肽(BNP)浓度及NICOM血流动力学参数,用logistic回归模型制作ROC曲线以评价NICOM参数和BNP对急性呼吸困难鉴别诊断的价值。结果AHF组BNP及胸腔液体含量较基线的变化率(TFCd0%)高于NAHF组,心指数(CI)低于NAHF组,差异有统计学意义(P<0.05)。logistic回归分析结果显示,BNP及TFCd0%是AHF独立危险因素(OR>1,P<0.05)。联合预测因子公式为L=BNP+327.5×TFCd0%,ROC曲线分析显示,联合预测因子L(AUC=93.7%)、BNP(AUC=89.9%)、TFCd0%(AUC=83.5%)及CI(AUC=63.4%)对AHF诊断均具有显著预测价值(P<0.05),并且联合预测因子L对AHF的诊断价值最高。结论NICOM监测TFCd0%和CI对AHF的诊断有较高的敏感度,但特异度较低,联合BNP检测可进一步提高其特异度。Objective To explore the clinical value of assessing the non-invasive cardiac output monitoring(NICOM)technology for differential diagnosis of acute dyspnea by applying receiver operating characteristic(ROC)curve.Methods A total of 80 patients with acute dyspnea admitted to the Department of Critical Care Medicine of Suzhou Ninth Hospital Affiliated to Soochow University from June 2019 to December 2021 were retrospectively analyzed.They were divided into a cardiac dyspnea(acute heart failure,AHF)group(n=38)and a non-cardiac dyspnea(non-AHF,NAHF)group(n=42)according to the etiology.The clinical data,concentration of serum B-type brain natriuretic peptide(BNP)and NICOM-related hemodynamic parameters were compared between the two groups.The values of NICOM parameters and BNP for the differential diagnosis of acute dyspnea were assessed by the ROC curve plotted with logistic regression models.Results The rates of changes of BNP and thoracic fluid content from baseline(TFCd0%)were higher,while the cardiac index(CI)was lower in the AHF group than in the NAHF group,with statistically significant differences(P<0.05).logistic regression analysis showed that BNP and TFCd0%were independent risk factors for AHF(OR>1,P<0.05).The combined predictor formula was L=BNP+327.5×TFCd0%.ROC curve analysis showed that the combined predictors L(AUC=93.7%),BNP(AUC=89.9%),TFCd0%(AUC=83.5%)and CI(AUC=63.4%)were of significant predictive value for the diagnosis of AHF(P<0.05),and the combined predictors L presented the highest diagnostic value for AHF.Conclusion NICOM technology for monitoring TFCd0%and CI has high sensitivity but low specificity for the diagnosis of AHF,and the combined BNP test can further improve the specificity.

关 键 词：无创血流动力学监测 生物电阻抗 B型脑钠肽 受试者工作曲线 呼吸困难 鉴别诊断 

分 类 号：R563.8[医药卫生—呼吸系统]