文献速递心脑血管急症复苏、休克与脓毒症重症综合管理与镇静镇痛

基于无监督机器学习的院外心脏骤停合并心肌梗死表型聚类

Phenotypic clustering of myocardial infarction complicated by out-of-hospital cardiac arrest using unsupervised machine learning.
PubMed 文献 · Resuscitation · 2026-03-05
一句话结论对478例心梗合并院外心脏骤停患者行k-means聚类,识别出3种表型,90天死亡率分别为22.5%、53.0%和77.2%。

该研究对478例因心肌梗死合并院外心脏骤停入院的前瞻性队列患者应用k-means无监督机器学习聚类,纳入人口学、心血管危险因素、心脏骤停特征、入院实验室数据、血流动力学参数及冠状动脉造影结果等变量。共识别出3个表型:第1类以年轻、合并症少、可电击心律为主、血流动力学良好为特征(n=260,54%);第2类以老年、高血压、糖尿病及既往冠心病为特征(n=118,25%);第3类以严重心源性休克、高乳酸、低凝血因子V、左室射血分数降低及频繁使用体外膜肺氧合为特征(n=100,21%)。第3类院内不良事件(全因死亡、出血、支架血栓)最频繁;90天死亡率第1类22.5%、第2类53.0%、第3类77.2%(p<0.001),与第1类相比,第2类和第3类90天死亡风险比分别为2.97(95%CI 2.07-4.26)和6.75(95%CI 4.74-9.60)。作者认为该表型分类可能有助于个体化管理与预后评估。

要点

  • 第3类(n=100,21%)以严重心源性休克、高乳酸、低凝血因子V、左室射血分数降低及频繁使用体外膜肺氧合为特征,院内全因死亡、出血、支架血栓最频繁。
  • 90天死亡率第1类22.5%、第2类53.0%、第3类77.2%(p<0.001),三组差异显著。
  • 与第1类相比,第2类90天死亡风险比2.97(95%CI 2.07-4.26),第3类6.75(95%CI 4.74-9.60)。
  • 第1类(n=260,54%)为年轻、合并症少、以可电击心律为主、血流动力学良好的患者,90天死亡率最低。
  • 第2类(n=118,25%)为老年、高血压、糖尿病及既往冠心病患者,90天死亡率居中。

适用人群与场景

  • 因心肌梗死合并院外心脏骤停入院的前瞻性队列患者,共478例。

研究设计与关键结果

  • 研究设计:对478例心肌梗死合并院外心脏骤停入院患者的前瞻性队列行k-means无监督机器学习聚类,候选变量包括人口学、心血管危险因素、心脏骤停特征、入院实验室数据、血流动力学参数及冠状动脉造影结果。
  • 识别出3个表型:第1类n=260(54%),年轻、合并症少、以可电击心律为主、血流动力学良好;第2类n=118(25%),老年、高血压、糖尿病、既往冠心病;第3类n=100(21%),严重心源性休克、高乳酸、低凝血因子V、左室射血分数降低、频繁使用体外膜肺氧合。
  • 院内不良事件(全因死亡、出血、支架血栓)在第3类最频繁。
  • 90天死亡率:第1类22.5%,第2类53.0%,第3类77.2%(p<0.001)。
  • 与第1类相比,90天死亡风险比:第2类2.97(95%CI 2.07-4.26),第3类6.75(95%CI 4.74-9.60)。

证据类型、等级与局限

  • 证据类型:基于前瞻性队列的无监督机器学习聚类分析。
  • 局限:原文未报告局限性。

对床旁实践的改变

  • 作者认为该表型分类可能有助于对该高危人群进行个体化管理与更精细的预后评估。

提炼说明

  • 本条依据论文摘要整理;摘要未涉及「诊断标准与关键阈值」「处置流程与药物剂量」「红旗征与不能漏的情况」「指南或共识出处」,故未列出这些栏目;全文请点「阅读原文」。
英文原文(摘要)· 点击展开

BACKGROUND: Patients with myocardial infarction (MI) complicated by out-of-hospital cardiac arrest (OHCA) represent a heterogeneous population with variable outcomes. Data-driven approaches may help uncover clinically meaningful subgroups to improve risk stratification and guide management.

METHODS: We applied an unsupervised machine learning analysis using k-means clustering to a prospective cohort of 478 patients admitted after OHCA related to MI. Candidate variables included demographics, cardiovascular risk factors, cardiac arrest characteristics, admission laboratory data, hemodynamic parameters, and coronary angiography findings. In-hospital outcomes included all-cause mortality, bleeding, and stent thrombosis. Ninety-day all-cause mortality was also assessed.

RESULTS: Three clusters were identified. Cluster 1 (n = 260, 54%) included younger patients with few comorbidities, predominantly shockable rhythms, and favorable hemodynamics. Cluster 2 (n = 118, 25%) included older patients with hypertension, diabetes, and prior coronary artery disease. Cluster 3 (n = 100, 21%) was characterized by severe cardiogenic shock, high lactate, low factor V, reduced left-ventricular ejection fraction, and frequent use of extracorporeal membrane oxygenation. Adverse in-hospital events, including all-cause mortality, bleeding, and stent thrombosis, were most frequent in cluster 3. Ninety-day mortality differed across groups: 22.5% in cluster 1, 53.0% in cluster 2, and 77.2% in cluster 3 (p < 0.001). Compared with cluster 1, hazard ratios for mortality at 90 days were 2.97 (95% CI, 2.07-4.26) in cluster 2 and 6.75 (95% CI: 4.74-9.60) in cluster 3.

CONCLUSIONS: Unsupervised machine learning identified three phenotypes among patients with MI-related OHCA associated with distinct outcomes. This phenotypic classification may facilitate personalized management and refined prognostic assessment in this high-risk population.

来源:Resuscitation,查看原文。版权归原作者所有。

参考来源

  1. Singh M, Cariou A, Varenne O, et al. Phenotypic clustering of myocardial infarction complicated by out-of-hospital cardiac arrest using unsupervised machine learning. Resuscitation. 2026. PMID: 41794115. doi:10.1016/j.resuscitation.2026.111037

供执业医师抢救与诊疗参考,不替代临床判断;剂量请结合体重、肝肾功能和本院药品规格复核,以指南原文为准。