Page 129 - 南京医科大学自然版
P. 129
第46卷第10期 姚运豪,陈旭锋,梅 勇. 基于多维度临床指标的体外心肺复苏患者30 d病死率预测模
2026年10月 型构建[J]. 南京医科大学学报(自然科学版),2026,46(10):1553-1560 ·1559 ·
模型面向 ECMO 上机存活 72 h 后的床旁预后分层, (5):163
目标为上机后治疗阶段风险评估,聚焦上机后脏器 [2] Supady A,Bělohlávek J,Combes A,et al. Extracorporeal
动态恢复指标,与时间指标分层互补,证实在多个 cardiopulmonary resuscitation for refractory cardiac
维度表现良好:AUC 为 0.836,校准曲线显示预测概 arrest[J]. Lancet Respir Med,2025,13(9):843-856
[3] 汤 丽,张春英,李君英,等. 血管活性药物评分在危重
率与实际观测高度吻合,DCA显示在宽阈值范围内
症患者中应用的研究进展[J]. 中华危重病急救医学,
具有临床净获益,Bootstrap 内部验证后一致性指数
2022,34(11):1213-1217
为 0.891,同时配套校准曲线、决策曲线分析三重评 Tang L,Zhang C Y,Li J Y,et al. Advances in the study of
估,充分证实模型在本中心人群内区分度、校准度及 vasoactive inotropic score in critically ill patients[J]. Chi⁃
临床净获益稳定,可满足院内风险分层使用需求,但 nese Critical Care Medicine,2022,34(11):1213-1217
在获得外部验证前,推广至其他中心仍需谨慎。 [4] Chen D L,Lin Y K,Li C I,et al. Impact of mechanical cir⁃
本研究存在若干局限性。首先,单中心回顾性 culatory support on out ⁃ of ⁃ hospital cardiac arrest out⁃
comes stratified by vasoactive⁃inotropic score:a retrospec⁃
设计样本量有限,可能存在选择偏倚;因国内ECPR
tive cohort study[J]. Resusc Plus,2024,19:100743
仍属于限制性急诊高级生命支持技术,仅少数大型
[5] Carlson J M,Etchill E,Whitman G,et al. Early withdraw⁃
三甲医院急诊中心可常年稳定开展,多中心 ECPR
al of life sustaining therapy in extracorporeal cardiopulmo⁃
标准化数据库尚未建立;后续计划联合多个具有稳
nary resuscitation(ECPR):results from the extracorporeal
定 ECPR 诊疗能力的三甲医院急诊中心,收集病例 life support organization registry[J]. Resuscitation,2022,
构建外部验证队列,以明确模型跨中心泛化效能。 179:71-77
其次,排除了ECMO上机后72 h内死亡病例,虽避免 [6] Dennis M,Supady A,Abrams D,et al. Recommendations
了早期技术性死亡的干扰,但也引入了“生存者偏 in extracorporeal cardiopulmonary resuscitation post re⁃
倚”。最后,未纳入院前急救时间指标(如低血流时 suscitation care via an international,modified delphi ap⁃
间、CPR至ECMO启动时间)及ECMO并发症等潜在 proach[J]. J Am Heart Assoc,2026,15(4):e047548
[7] El⁃Menyar A,Wahlen B M. Cardiac arrest,stony heart,
混杂因素,后续前瞻性研究将规范院前急救数据采
and cardiopulmonary resuscitation:an updated revisit[J].
集,完整记录骤停时间相关参数,将其纳入模型进
World J Cardiol,2024,16(3):126-136
一步优化校正混杂。
[8] Jeppesen K K,Rasmussen S B,Kjaergaard J,et al. Acute
综上所述,针对APACHE Ⅱ评分、CHD、CRRT、 kidney injury after out⁃of⁃hospital cardiac arrest[J]. Crit
ΔcTnT 及 VISmax72 h 5 项指标构建的预测模型,可 Care,2024,28(1):169
用于存活超过72 h的成人ECPR患者30 d死亡风险 [9] Li C,Wang Y,Wang X,et al. Acute kidney injury and car⁃
分层,具备一定临床应用价值。 diogenic shock severity for mortality risk stratification in
利益冲突声明: patients supported with VA ECMO[J]. ESC Heart Fail,
所有作者均声明不存在利益冲突。 2024,11(6):3872-3881
Conflict of Interests: [10]Fleming G M,Askenazi D J,Bridges B C,et al. A multi⁃
center international survey of renal supportive therapy
All authors declare that there is no conflict of interests.
作者贡献声明: during ECMO:the kidney intervention during extracorpo⁃
real membrane oxygenation(KIDMO)group[J]. ASAIO J,
姚运豪负责研究设计、数据收集与分析、作图及论文撰
写;梅勇负责研究指导、论文审阅;陈旭锋负责研究指导、论 2012,58(4):407-414
文审阅。 [11]Hoste E A,Bagshaw S M,Bellomo R,et al. Epidemiology
of acute kidney injury in critically ill patients:the multi⁃
Author’s Contributions:
national AKI⁃EPI study[J]. Intensive Care Med,2015,41
Yao Yunhao was responsible for study design,data collec⁃
(8):1411-1423
tion and analysis,plotting,and manuscript writing;Mei Yong
[12]Surjit A,Prasannan B,Abraham J,et al. Acute kidney in⁃
was responsible for research guidance and manuscript review⁃
jury in patients undergoing extracorporeal membrane oxy⁃
ing;Chen Xufeng was responsible for research guidance and
genation:a retrospective cohort study[J]. Indian J Crit
manuscript reviewing.
Care Med,2024,28(1):26-29
[参考文献]
[13]Zhang D,Li L,Huang W,et al. Vasoactive⁃inotropic score as
[1] Klick R. Extracorporeal cardiopulmonary resuscitation⁃ev⁃ a promising predictor of acute kidney injury in adult pa⁃
idence and implications[J]. Dtsch Arztebl Int,2024,121 tients requiring extracorporeal membrane oxygenation[J].

