文献速递呼吸衰竭与机械通气重症综合管理与镇静镇痛

机器学习分类模型可利用常规临床数据识别急性呼吸窘迫综合征表型

Machine Learning Classifier Models Can Identify Acute Respiratory Distress Syndrome Phenotypes Using Readily Available Clinical Data.
PubMed 文献 · Am J Respir Crit Care Med · 2020-01-01
一句话结论基于24项常规临床数据的梯度提升机器学习模型可准确识别ARDS高炎症与低炎症表型(验证队列AUC 0.95),且模型分类的表型与治疗交互作用显著。

该研究利用三项随机对照试验队列(ARMA、ALVEOLI、FACTT,n=2,022)作为训练集,以SAILS队列(n=745)为验证集,采用梯度提升机器学习算法基于入组时24项人口学、生命体征、实验室和呼吸变量构建ARDS表型分类模型。主要分析中模型在验证队列准确分类表型(AUC 0.95,95%CI 0.94-0.96);以0.5概率截断分类时,模型判定的高炎症表型患者炎症生物标志物(IL-6、IL-8、sTNFR-1,P<0.0001)和90天死亡率(38% vs 24%,P=0.0002)显著更高。以ALVEOLI(AUC 0.94,95%CI 0.92-0.96)和FACTT(AUC 0.94,95%CI 0.92-0.95)分别作为验证队列时模型准确度相似。在ALVEOLI(P=0.0113)和FACTT(P=0.0072)队列中观察到临床分类器模型判定表型与治疗的显著交互作用。结论认为基于常规临床数据的机器学习模型可准确识别ARDS表型,可能有助于床旁快速表型识别。

要点

  • 模型仅使用入组时24项常规临床数据(人口学、生命体征、实验室、呼吸变量)即可分类ARDS表型,验证队列AUC 0.95(95%CI 0.94-0.96)。
  • 以概率截断0.5分类时,模型判定的高炎症表型患者90天死亡率38% vs 低炎症表型24%(P=0.0002),炎症生物标志物IL-6、IL-8、sTNFR-1亦显著更高(P<0.0001)。
  • 以ALVEOLI和FACTT分别作为验证队列时模型准确度相似(AUC均为0.94,95%CI分别为0.92-0.96和0.92-0.95)。
  • 在ALVEOLI(P=0.0113)和FACTT(P=0.0072)队列中,模型判定的表型与随机治疗之间存在显著交互作用。
  • 模型有望实现床旁快速表型识别,但需进一步研究验证。

适用人群与场景

  • 三项随机对照试验队列(ARMA、ALVEOLI、FACTT)的ARDS患者作为训练集,共n=2,022。
  • SAILS随机对照试验队列的ARDS患者作为验证集,n=745。

诊断标准与关键阈值

  • 模型输入变量为入组时24项常规临床数据,包括人口学、生命体征、实验室和呼吸变量。
  • 以概率截断值0.5将患者分类为高炎症表型或低炎症表型。
  • 模型性能以潜类别分析衍生的表型为参照标准进行评价。

研究设计与关键结果

  • 研究为基于随机对照试验队列的模型开发与验证研究,训练集n=2,022(ARMA、ALVEOLI、FACTT),验证集n=745(SAILS)。
  • 主要分析中模型在验证队列准确分类ARDS表型,AUC 0.95(95%CI 0.94-0.96)。
  • 以0.5概率截断分类时,模型判定的高炎症表型患者炎症生物标志物(IL-6、IL-8、sTNFR-1)显著高于低炎症表型(P<0.0001)。
  • 模型判定的高炎症表型患者90天死亡率38% vs 低炎症表型24%(P=0.0002)。
  • 以ALVEOLI为验证队列时模型AUC 0.94(95%CI 0.92-0.96),以FACTT为验证队列时AUC 0.94(95%CI 0.92-0.95)。
  • 在ALVEOLI队列中观察到模型判定表型与治疗的显著交互作用(P=0.0113),在FACTT队列中亦观察到显著交互作用(P=0.0072)。

证据类型、等级与局限

  • 证据类型:基于三项随机对照试验队列训练、一项随机对照试验队列验证的模型开发与验证研究。
  • 局限:原文未报告局限性。

对床旁实践的改变

  • 作者认为基于常规临床数据的机器学习模型可准确识别ARDS表型,可能有助于床旁快速表型识别。
  • 该模型使用入组时24项常规临床数据,无需血浆生物标志物即可分类表型,可能克服当前缺乏床旁检测的限制。

提炼说明

  • 本条依据论文摘要整理;摘要未涉及「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

Rationale: Two distinct phenotypes of acute respiratory distress syndrome (ARDS) with differential clinical outcomes and responses to randomly assigned treatment have consistently been identified in randomized controlled trial cohorts using latent class analysis. Plasma biomarkers, key components in phenotype identification, currently lack point-of-care assays and represent a barrier to the clinical implementation of phenotypes.Objectives: The objective of this study was to develop models to classify ARDS phenotypes using readily available clinical data only.Methods: Three randomized controlled trial cohorts served as the training data set (ARMA [High vs. Low Vt], ALVEOLI [Assessment of Low Vt and Elevated End-Expiratory Pressure to Obviate Lung Injury], and FACTT [Fluids and Catheter Treatment Trial]; n = 2,022), and a fourth served as the validation data set (SAILS [Statins for Acutely Injured Lungs from Sepsis]; n = 745). A gradient-boosted machine algorithm was used to develop classifier models using 24 variables (demographics, vital signs, laboratory, and respiratory variables) at enrollment. In two secondary analyses, the ALVEOLI and FACTT cohorts each, individually, served as the validation data set, and the remaining combined cohorts formed the training data set for each analysis. Model performance was evaluated against the latent class analysis-derived phenotype.Measurements and Main Results: For the primary analysis, the model accurately classified the phenotypes in the validation cohort (area under the receiver operating characteristic curve [AUC], 0.95; 95% confidence interval [CI], 0.94-0.96). Using a probability cutoff of 0.5 to assign class, inflammatory biomarkers (IL-6, IL-8, and sTNFR-1; P < 0.0001) and 90-day mortality (38% vs. 24%; P = 0.0002) were significantly higher in the hyperinflammatory phenotype as classified by the model. Model accuracy was similar when ALVEOLI (AUC, 0.94; 95% CI, 0.92-0.96) and FACTT (AUC, 0.94; 95% CI, 0.92-0.95) were used as the validation cohorts. Significant treatment interactions were observed with the clinical classifier model-assigned phenotypes in both ALVEOLI (P = 0.0113) and FACTT (P = 0.0072) cohorts.Conclusions: ARDS phenotypes can be accurately identified using machine learning models based on readily available clinical data and may enable rapid phenotype identification at the bedside.

来源:Am J Respir Crit Care Med,查看原文。版权归原作者所有。

参考来源

  1. Sinha P, Churpek MM, Calfee CS, et al. Machine Learning Classifier Models Can Identify Acute Respiratory Distress Syndrome Phenotypes Using Readily Available Clinical Data. Am J Respir Crit Care Med. 2020. PMID: 32551817. doi:10.1164/rccm.202002-0347OC

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。