Department of Emergency Medicine, Massachusetts General Hospital, Harvard Medical School, Boston.
Division of Emergency Medicine, Children's National Health System, Washington, DC.
JAMA Netw Open. 2019 Jan 4;2(1):e186937. doi: 10.1001/jamanetworkopen.2018.6937.
While machine learning approaches may enhance prediction ability, little is known about their utility in emergency department (ED) triage.
To examine the performance of machine learning approaches to predict clinical outcomes and disposition in children in the ED and to compare their performance with conventional triage approaches.
DESIGN, SETTING, AND PARTICIPANTS: Prognostic study of ED data from the National Hospital Ambulatory Medical Care Survey from January 1, 2007, through December 31, 2015. A nationally representative sample of 52 037 children aged 18 years or younger who presented to the ED were included. Data analysis was performed in August 2018.
The outcomes were critical care (admission to an intensive care unit and/or in-hospital death) and hospitalization (direct hospital admission or transfer). In the training set (70% random sample), using routinely available triage data as predictors (eg, demographic characteristics and vital signs), we derived 4 machine learning-based models: lasso regression, random forest, gradient-boosted decision tree, and deep neural network. In the test set (the remaining 30% of the sample), we measured the models' prediction performance by computing C statistics, prospective prediction results, and decision curves. These machine learning models were built for each outcome and compared with the reference model using the conventional triage classification information.
Of 52 037 eligible ED visits by children (median [interquartile range] age, 6 [2-14] years; 24 929 [48.0%] female), 163 (0.3%) had the critical care outcome and 2352 (4.5%) had the hospitalization outcome. For the critical care prediction, all machine learning approaches had higher discriminative ability compared with the reference model, although the difference was not statistically significant (eg, C statistics of 0.85 [95% CI, 0.78-0.92] for the deep neural network vs 0.78 [95% CI, 0.71-0.85] for the reference; P = .16), and lower number of undertriaged critically ill children in the conventional triage levels 3 to 5 (urgent to nonurgent). For the hospitalization prediction, all machine learning approaches had significantly higher discrimination ability (eg, C statistic, 0.80 [95% CI, 0.78-0.81] for the deep neural network vs 0.73 [95% CI, 0.71-0.75] for the reference; P < .001) and fewer overtriaged children who did not require inpatient management in the conventional triage levels 1 to 3 (immediate to urgent). The decision curve analysis demonstrated a greater net benefit of machine learning models over ranges of clinical thresholds.
Machine learning-based triage had better discrimination ability to predict clinical outcomes and disposition, with reduction in undertriaging critically ill children and overtriaging children who are less ill.
虽然机器学习方法可能会提高预测能力，但人们对其在急诊科（ED）分诊中的应用知之甚少。
研究机器学习方法在预测 ED 中儿童临床结局和处置方面的表现，并比较其与传统分诊方法的表现。
设计、设置和参与者：这是一项来自 2007 年 1 月 1 日至 2015 年 12 月 31 日全国医院门诊医疗调查的 ED 数据的预后研究。研究纳入了年龄在 18 岁或以下、52037 名来自全国代表性样本的儿童。数据分析于 2018 年 8 月进行。
结局为重症监护（入住重症监护病房和/或院内死亡）和住院（直接住院或转院）。在训练集中（70%的随机样本），我们使用常规分诊数据作为预测因素（例如，人口统计学特征和生命体征），推导出 4 种基于机器学习的模型：套索回归、随机森林、梯度提升决策树和深度神经网络。在测试集中（剩余的 30%样本），我们通过计算 C 统计量、前瞻性预测结果和决策曲线来衡量模型的预测性能。为每个结局构建了这些机器学习模型，并与使用传统分诊分类信息的参考模型进行了比较。
在 52037 名符合条件的儿童 ED 就诊中（中位数[四分位距]年龄，6[2-14]岁；24929[48.0%]为女性），有 163 名（0.3%）出现了危急重症结局，2352 名（4.5%）出现了住院结局。对于危急重症预测，所有机器学习方法与参考模型相比，均具有更高的区分能力，尽管差异无统计学意义（例如，深度神经网络的 C 统计量为 0.85[95%CI，0.78-0.92]，参考模型为 0.78[95%CI，0.71-0.85]；P=0.16），且传统分诊级别 3-5（紧急至非紧急）中危急重症儿童的分诊不足人数更少。对于住院预测，所有机器学习方法的判别能力均显著提高（例如，深度神经网络的 C 统计量为 0.80[95%CI，0.78-0.81]，参考模型为 0.73[95%CI，0.71-0.75]；P<0.001），且传统分诊级别 1-3（立即至紧急）中不需要住院管理的过度分诊儿童人数减少。决策曲线分析表明，机器学习模型在临床阈值范围内具有更大的净收益。
基于机器学习的分诊具有更好的预测临床结局和处置的区分能力，可以减少危急重症儿童的分诊不足和病情较轻的儿童的过度分诊。
