Predicting Risk of Suicide Attempts Over Time Through Machine Learning

被引:366
作者
Walsh, Colin G. [1 ,2 ,3 ]
Ribeiro, Jessica D. [4 ]
Franklin, Joseph C. [4 ]
机构
[1] Vanderbilt Univ, Med Ctr, Dept Biomed Informat, 2525 West End Ave,Ste 1475, Nashville, TN 37203 USA
[2] Vanderbilt Univ, Med Ctr, Dept Med, 2525 West End Ave,Ste 1475, Nashville, TN 37203 USA
[3] Vanderbilt Univ, Med Ctr, Dept Psychiat, 2525 West End Ave,Ste 1475, Nashville, TN 37203 USA
[4] Florida State Univ, Dept Psychol, Tallahassee, FL 32306 USA
关键词
suicide prevention; prediction; prevention; classification; CLASSIFICATION; METAANALYSIS; MORTALITY; MODELS; DISORDERS; BEHAVIORS; THOUGHTS; SYSTEM; TREES;
D O I
10.1177/2167702617691560
中图分类号
B849 [应用心理学];
学科分类号
040203 ;
摘要
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.
引用
收藏
页码:457 / 469
页数:13
相关论文
共 42 条
[1]   A prediction model for COPD readmissions: catching up, catching our breath, and improving a national problem [J].
Amalakuhan, Bravein ;
Kiljanek, Lukasz ;
Parvathaneni, Arvin ;
Hester, Michael ;
Cheriyath, Pramil ;
Fischman, Daniel .
JOURNAL OF COMMUNITY HOSPITAL INTERNAL MEDICINE PERSPECTIVES, 2012, 2 (01)
[2]  
[Anonymous], 2013, SAS GLOBAL FORUM 201
[3]  
[Anonymous], RANGER FAST IMPLEMTA
[4]  
[Anonymous], 2016, SUIC DAT
[5]  
[Anonymous], 2011, J. Mach. Learn. Res.
[6]  
[Anonymous], HLTH CARE FINANCE RE
[7]  
[Anonymous], 2016, INJ PREV CONTR DAT S
[8]  
[Anonymous], PREDICTING SUICIDES
[9]  
[Anonymous], 2012, R LANG ENV STAT COMP
[10]  
[Anonymous], 2004, MULTIPLE IMPUTATION