Predicting Risk of Suicide Attempts Over Time Through Machine Learning
Predicting Risk of Suicide Attempts Over Time Through Machine Learning
复制标题
DOI：
10.1177/2167702617691560
复制
发表时间：
2017-05-01
影响因子：
4.8
通讯作者：
Franklin, Joseph C.
中科院分区：
文献类型：
--
作者：
Walsh, Colin G.;Ribeiro, Jessica D.;Franklin, Joseph C.
预测自杀未遂的传统方法限制了这些危险行为风险检测的准确性和规模。我们试图通过将机器学习应用于大型医疗数据库中的电子健康记录来克服这些限制。参与者是 5,167 名具有自伤索赔代码（即 ICD-9、E95x）的成年患者；专家对记录的审查确定，有 3,250 名患者尝试自杀（即病例），1,917 名患者进行了非自杀、意外或无法证实的自残行为（即对照）。我们开发了机器学习算法，可以准确预测未来的自杀企图（AUC = 0.84，精度 = 0.79，召回率 = 0.95，Brier 得分 = 0.14）。此外，准确性从自杀企图前 720 天提高到 7 天，并且预测因子的重要性随着时间的推移而变化。这些发现代表了朝着准确和可扩展的风险检测迈出的一步，并提供了有关自杀未遂风险如何随时间变化的见解。
Traditional approaches to the prediction of suicide attempts have limited the accuracy and scale of risk detection for these dangerous behaviors. We sought to overcome these limitations by applying machine learning to electronic health records within a large medical database. Participants were 5,167 adult patients with a claim code for self-injury (i.e., ICD-9, E95x); expert review of records determined that 3,250 patients made a suicide attempt (i.e., cases), and 1,917 patients engaged in self-injury that was nonsuicidal, accidental, or nonverifiable (i.e., controls). We developed machine learning algorithms that accurately predicted future suicide attempts (AUC = 0.84, precision = 0.79, recall = 0.95, Brier score = 0.14). Moreover, accuracy improved from 720 days to 7 days before the suicide attempt, and predictor importance shifted across time. These findings represent a step toward accurate and scalable risk detection and provide insight into how suicide attempt risk shifts over time.