文献速递心脑血管急症重症综合管理与镇静镇痛

机器学习识别脑出血后患者结局的可改变预测因素

Identifying Modifiable Predictors of Patient Outcomes After Intracerebral Hemorrhage with Machine Learning.
PubMed 文献 · Neurocrit Care · 2021-01-01
一句话结论在两项前瞻性队列中,血肿扩大是脑出血后功能结局唯一可改变独立预测因素;格拉斯哥昏迷量表评分高、入院血肿体积<44.5 mL且病前mRS较低者可能为可改变结局人群。

研究前瞻性纳入两家机构的脑出血患者,收集血肿体积、用药史等变量,以14天和3个月改良Rankin量表(mRS)0-3分为良好结局、4-6分为不良结局。在机构1训练监督机器学习模型,经重复五折交叉验证并在机构2独立样本测试,用ROC曲线下面积(AUC)评估拟合。机构1随机森林AUC为14天0.85(0.84,0.87)、3个月0.82(0.80,0.84);机构2决策树AUC 0.69(0.63,0.75)、随机森林0.78(0.72,0.84)。算法选出的重要独立预测因素包括入院血肿体积、血肿扩大、脑室出血、总体ICH评分和格拉斯哥昏迷量表评分,其中血肿扩大是唯一可改变独立预测因素。

要点

  • 血肿扩大是脑出血后14天和3个月功能结局唯一可改变的独立预测因素,跨两种结局类型和建模方法一致。
  • 3个月结局决策树分析显示,格拉斯哥昏迷量表评分高、入院血肿体积<44.5 mL且病前mRS相对低的患者具有可改变结局,似乎为未来干预改善结局的候选人群。
  • 机构1随机森林预测14天良好vs不良结局的AUC为0.85(0.84,0.87),预测3个月结局的AUC为0.82(0.80,0.84)。
  • 在机构2独立样本中,决策树AUC为0.69(0.63,0.75),随机森林AUC为0.78(0.72,0.84),模型预测良好。
  • 算法选出的重要独立预测因素包括入院血肿体积、血肿扩大、脑室出血、总体ICH评分和格拉斯哥昏迷量表评分。

适用人群与场景

  • 在两个机构前瞻性识别的脑出血患者。
  • 结局评估为14天和3个月改良Rankin量表(mRS),良好结局定义为0-3分(独立或更好),不良结局定义为4-6分(依赖或更差)。

诊断标准与关键阈值

  • 良好结局:改良Rankin量表(mRS)0-3分(独立或更好);不良结局:mRS 4-6分(依赖或更差)。
  • 决策树分析预测3个月结局时,格拉斯哥昏迷量表评分高、入院血肿体积<44.5 mL且病前mRS相对低的患者具有可改变结局。

研究设计与关键结果

  • 研究类型为前瞻性识别患者的两机构研究,使用监督机器学习模型识别良好vs不良结局的最佳预测因素,经重复五折交叉验证并在机构2独立样本测试,以ROC曲线下面积(AUC)评估模型拟合。
  • 机构1模型性能:14天决策树AUC 0.79(0.77,0.81),随机森林AUC 0.85(0.84,0.87);3个月决策树AUC 0.75(0.73,0.77),随机森林AUC 0.82(0.80,0.84)。
  • 算法选出的重要独立预测因素包括入院血肿体积、血肿扩大、脑室出血、总体ICH评分和格拉斯哥昏迷量表评分。
  • 血肿扩大是唯一可改变的独立预测因素,在部分入院血肿体积低、格拉斯哥昏迷量表评分好和病前mRS评分低的患者中,与良好或不良结局均相容。
  • 在机构2使用协调数据训练的模型预测结局:决策树AUC 0.69(0.63,0.75),随机森林AUC 0.78(0.72,0.84)。

证据类型、等级与局限

  • 证据类型:前瞻性识别患者的两机构研究,使用监督机器学习建模。
  • 局限:原文未报告局限性。

对床旁实践的改变

  • 作者认为,根据预测3个月结局的决策树分析,格拉斯哥昏迷量表评分高、入院血肿体积<44.5 mL且病前mRS相对低的患者具有可改变结局,似乎为未来干预改善脑出血后结局的候选人群。

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

BACKGROUND/OBJECTIVE: Demonstrating a benefit of acute treatment to patients with intracerebral hemorrhage (ICH) requires identifying which patients have a potentially modifiable outcome, where treatment could favorably shift a patient's expected outcome. A decision rule for which patients have a modifiable outcome could improve the targeting of treatments. We sought to determine which patients with ICH have a modifiable outcome.

METHODS: Patients with ICH were prospectively identified at two institutions. Data on hematoma volumes, medication histories, and other variables of interest were collected. ICH outcomes were evaluated using the modified Rankin Scale (mRS), assessed at 14 days and 3 months after ICH, with "good outcome" defined as 0-3 (independence or better) and "poor outcome" defined as 4-6 (dependence or worse). Supervised machine learning models identified the best predictors of good versus poor outcomes at Institution 1. Models were validated using repeated fivefold cross-validation as well as testing on the entirely independent sample at Institution 2. Model fit was assessed with area under the ROC curve (AUC).

RESULTS: Model performance at Institution 1 was strong for both 14-day (AUC of 0.79 [0.77, 0.81] for decision tree, 0.85 [0.84, 0.87] for random forest) and 3 month (AUC of 0.75 [0.73, 0.77] for decision tree, 0.82 [0.80, 0.84] for random forest) outcomes. Independent predictors of functional outcome selected by the algorithms as important included hematoma volume at hospital admission, hematoma expansion, intraventricular hemorrhage, overall ICH Score, and Glasgow Coma Scale. Hematoma expansion was the only potentially modifiable independent predictor of outcome and was compatible with "good" or "poor" outcome in a subset of patients with low hematoma volumes, good Glasgow Coma scale and premorbid modified Rankin Scale scores. Models trained on harmonized data also predicted patient outcomes well at Institution 2 using decision tree (AUC 0.69 [0.63, 0.75]) and random forests (AUC 0.78 [0.72, 0.84]).

CONCLUSIONS: Patient outcomes are predictable to a high level in patients with ICH, and hematoma expansion is the sole-modifiable predictor of these outcomes across two outcome types and modeling approaches. According to decision tree analyses predicting outcome at 3 months, patients with a high Glasgow Coma Scale score, less than 44.5 mL hematoma volume at admission, and relatively low premorbid modified Rankin Score in particular have a modifiable outcome and appear to be candidates for future interventions to improve outcomes after ICH.

来源:Neurocrit Care,查看原文。版权归原作者所有。

参考来源

  1. Hall AN, Weaver B, Liotta E, et al. Identifying Modifiable Predictors of Patient Outcomes After Intracerebral Hemorrhage with Machine Learning. Neurocrit Care. 2021. PMID: 32385834. doi:10.1007/s12028-020-00982-8

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。