Development of an early-warning system for high-risk patients for suicide attempt using deep learning and electronic health records

Suicide is the tenth leading cause of death in the United States (US). An early-warning system (EWS) for suicide attempt could prove valuable for identifying those at risk of suicide attempts, and analyzing the contribution of repeated attempts to the risk of eventual death by suicide. In this study we sought to develop an EWS for high-risk suicide attempt patients through the development of a population-based risk stratification surveillance system. Advanced machine-learning algorithms and deep neural networks were utilized to build models with the data from electronic health records (EHRs). A final risk score was calculated for each individual and calibrated to indicate the probability of a suicide attempt in the following 1-year time period. Risk scores were subjected to individual-level analysis in order to aid in the interpretation of the results for health-care providers managing the at-risk cohorts. The 1-year suicide attempt risk model attained an area under the curve (AUC ROC) of 0.792 and 0.769 in the retrospective and prospective cohorts, respectively. The suicide attempt rate in the “very high risk” category was 60 times greater than the population baseline when tested in the prospective cohorts. Mental health disorders including depression, bipolar disorders and anxiety, along with substance abuse, impulse control disorders, clinical utilization indicators, and socioeconomic determinants were recognized as significant features associated with incident suicide attempt.

[1]  R. C. Macridis A review , 1963 .

[2]  M. Pagano,et al.  Survival analysis. , 1996, Nutrition.

[3]  J. Lönnqvist,et al.  Suicide attempts preceding completed suicide , 1998, British Journal of Psychiatry.

[4]  J. Kruse,et al.  Application of stratum-specific likelihood ratios in mental health screening , 2000, Social Psychiatry and Psychiatric Epidemiology.

[5]  S. Buka,et al.  The association between suicide and the socio-economic characteristics of geographical areas: a systematic review , 2005, Psychological Medicine.

[6]  D. Wennberg,et al.  Case finding for patients at risk of readmission to hospital: development of algorithm to identify high risk patients , 2006, BMJ : British Medical Journal.

[7]  M. Fushimi,et al.  Comparison of completed and attempted suicide in Akita, Japan , 2006, Psychiatry and clinical neurosciences.

[8]  Sharon Friel,et al.  Closing the gap in a generation: health equity through action on the social determinants of health , 2008, The Lancet.

[9]  A. Haas,et al.  The American Foundation for Suicide Prevention , 2009 .

[10]  J. Overholser,et al.  Apples to oranges?: a direct comparison between suicide attempters and suicide completers. , 2010, Journal of affective disorders.

[11]  V. Preedy,et al.  Prospective Cohort Study , 2010 .

[12]  K. Kushner,et al.  'Health equity through action on the social determinants of health': taking up the challenge in nursing. , 2010, Nursing inquiry.

[13]  David R. Williams,et al.  Twelve-month prevalence of and risk factors for suicide attempts in the World Health Organization World Mental Health Surveys. , 2010, The Journal of clinical psychiatry.

[14]  K. Merikangas,et al.  Prevalence and correlates of eating disorders in adolescents. Results from the national comorbidity survey replication adolescent supplement. , 2011, Archives of general psychiatry.

[15]  J. Gfroerer,et al.  Suicidal thoughts and behaviors among adults aged ≥18 years--United States, 2008-2009. , 2011, Morbidity and mortality weekly report. Surveillance summaries.

[16]  Preben Bo Mortensen,et al.  Absolute risk of suicide after first hospital contact in mental disorder. , 2011, Archives of general psychiatry.

[17]  E. Mohammadi,et al.  Barriers and facilitators related to the implementation of a physiological track and trigger system: A systematic review of the qualitative evidence , 2017, International journal for quality in health care : journal of the International Society for Quality in Health Care.

[18]  Matthew K Nock,et al.  Prevalence, correlates, and treatment of lifetime suicidal behavior among adolescents: results from the National Comorbidity Survey Replication Adolescent Supplement. , 2013, JAMA psychiatry.

[19]  T. Terao,et al.  Risk factors for suicide in Japan: a model of predicting suicide in 2008 by risk factors of 2007. , 2013, Journal of affective disorders.

[20]  J. Stine,et al.  Hepatotoxicity of antibiotics: a review and update for the clinician. , 2013, Clinics in liver disease.

[21]  I. Katz,et al.  Noncancer pain conditions and risk of suicide. , 2013, JAMA psychiatry.

[22]  I. Katz,et al.  Predictive Modeling and Concentration of the Risk of Suicide: Implications for Preventive Interventions in the US Department of Veterans Affairs. , 2015, American journal of public health.

[23]  L. S. Too,et al.  Interventions to reduce suicides at suicide hotspots: a systematic review and meta-analysis. , 2015, The lancet. Psychiatry.

[24]  R. Ries,et al.  Addiction and suicide: A review. , 2015, The American journal on addictions.

[25]  A. Friedman Maternal early warning systems. , 2015, Obstetrics and gynecology clinics of North America.

[26]  Wei Zhang,et al.  Does disability predict attempted suicide in the elderly? A community-based study of elderly residents in Shanghai, China , 2016, Aging & mental health.

[27]  P. Lichtenstein,et al.  Suicide risk after nonfatal self-harm: a national cohort study, 2000-2008. , 2016, The Journal of clinical psychiatry.

[28]  Tianqi Chen,et al.  XGBoost: A Scalable Tree Boosting System , 2016, KDD.

[29]  D. Osborn,et al.  Bereavement by suicide as a risk factor for suicide attempt: a cross-sectional national UK-wide study of 3432 young bereaved adults , 2016, BMJ Open.

[30]  Sameer Singh,et al.  “Why Should I Trust You?”: Explaining the Predictions of Any Classifier , 2016, NAACL.

[31]  Svetha Venkatesh,et al.  Predicting Risk of Suicide Attempt Using History of Physical Illnesses From Electronic Medical Records , 2016, JMIR mental health.

[32]  Carlos Guestrin,et al.  "Why Should I Trust You?": Explaining the Predictions of Any Classifier , 2016, ArXiv.

[33]  D. de Leo,et al.  Suicide in older adults: a comparison with middle-aged adults using the Queensland Suicide Register , 2016, International Psychogeriatrics.

[34]  J. Bostwick,et al.  Suicide Attempt as a Risk Factor for Completed Suicide: Even More Lethal Than We Knew. , 2016, The American journal of psychiatry.

[35]  Camille Locklear Goins,et al.  A Phenomenological Study , 2008 .

[36]  R. Weisler,et al.  Suicide Behaviors in Bipolar Disorder: A Review and Update for the Clinician. , 2016, The Psychiatric clinics of North America.

[37]  T. Terao,et al.  Anxious temperament as a risk factor of suicide attempt. , 2016, Comprehensive psychiatry.

[38]  F. Bellivier,et al.  Serious suicide attempts in outpatients with multiple substance use disorders. , 2017, Drug and alcohol dependence.

[39]  D. D. De Beurs,et al.  Network Analysis: A Novel Approach to Understand Suicidal Behaviour , 2017, International Journal of Environmental Research and Public Health.

[40]  Shoab Saadat,et al.  Predicting Quality of Life Changes in Hemodialysis Patients Using Machine Learning: Generation of an Early Warning System , 2017, Cureus.

[41]  Diego J. Palao-Vidal,et al.  Risk of re-attempts and suicide death after a suicide attempt: A survival analysis , 2017, BMC Psychiatry.

[42]  Norio Sugawara,et al.  Application of a stratum-specific likelihood ratio analysis in a screen for depression among a community-dwelling population in Japan , 2017, Neuropsychiatric disease and treatment.

[43]  P. A. Bradley,et al.  Developing a practical suicide risk prediction model for targeting high‐risk patients in the Veterans health Administration , 2017, International journal of methods in psychiatric research.

[44]  Matthew K Nock,et al.  Predicting Suicidal Behavior From Longitudinal Electronic Health Records. , 2017, The American journal of psychiatry.

[45]  J. Draaisma,et al.  Clinician Perceptions of an Early Warning System on Patient Safety. , 2017, Hospital pediatrics.

[46]  David C. Atkins,et al.  Prediction of Suicide Ideation and Attempt Among Substance-Using Patients in Primary Care , 2017, The Journal of the American Board of Family Medicine.

[47]  Bo Jin,et al.  Estimating One-Year Risk of Incident Chronic Kidney Disease: Retrospective Development and Validation Study Using Electronic Medical Record Data From the State of Maine , 2017, JMIR medical informatics.

[48]  A. Mcgrady,et al.  Psychosocial Factors and Comorbidity Associated with Suicide Attempts: Findings in Patients with Bipolar Disorder , 2017, Psychopathology.

[49]  C. Bambra,et al.  Area-level socioeconomic disadvantage and suicidal behaviour in Europe: A systematic review. , 2017, Social science & medicine.

[50]  D. Beurs Network Analysis: A Novel Approach to Understand Suicidal Behaviour , 2017 .

[51]  Colin G. Walsh,et al.  Predicting Risk of Suicide Attempts Over Time Through Machine Learning , 2017 .

[52]  E. Arias,et al.  Deaths: Final Data for 2015. , 2017, National vital statistics reports : from the Centers for Disease Control and Prevention, National Center for Health Statistics, National Vital Statistics System.

[53]  Evan M. Kleiman,et al.  Risk Factors for Suicidal Thoughts and Behaviors: A Meta-Analysis of 50 Years of Research , 2017, Psychological bulletin.

[54]  J. Mann,et al.  Are comorbid anxiety disorders a risk factor for suicide attempts in patients with mood disorders? A two-year prospective study , 2018, European Psychiatry.

[55]  M. Arns,et al.  Predicting sex from brain rhythms with deep learning , 2018, Scientific Reports.

[56]  Byoung-Tak Zhang,et al.  Deep ECGNet: An Optimal Deep Learning Framework for Monitoring Mental Stress Using Ultra Short-Term ECG Signals. , 2018, Telemedicine journal and e-health : the official journal of the American Telemedicine Association.

[57]  S. Siedlecki,et al.  Enhanced early warning system impact on nursing practice: A phenomenological study , 2018, Journal of advanced nursing.

[58]  P. Seed,et al.  Early warning system hypertension thresholds to predict adverse outcomes in pre-eclampsia: A prospective cohort study , 2017, Pregnancy hypertension.

[59]  Hyeongrae Lee,et al.  Use of a Machine Learning Algorithm to Predict Individuals with Suicide Ideation in the General Population , 2018, Psychiatry investigation.

[60]  H. Hedegaard,et al.  Issues in Developing a Surveillance Case Definition for Nonfatal Suicide Attempt and Intentional Self-harm Using International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) Coded Data. , 2018, National health statistics reports.

[61]  M. Delgado-Rodríguez,et al.  Systematic review and meta-analysis. , 2017, Medicina intensiva.

[62]  Thomas Blaschke,et al.  The rise of deep learning in drug discovery. , 2018, Drug discovery today.

[63]  HwangBosun,et al.  Deep ECGNet: An Optimal Deep Learning Framework for Monitoring Mental Stress Using Ultra Short-Term ECG Signals. , 2018 .

[64]  Simo Saarakkala,et al.  Automatic Knee Osteoarthritis Diagnosis from Plain Radiographs: A Deep Learning-Based Approach , 2017, Scientific Reports.

[65]  Colin G. Walsh,et al.  Predicting suicide attempts in adolescents with longitudinal clinical data and machine learning , 2018, Journal of child psychology and psychiatry, and allied disciplines.

[66]  E. Krupinski Deep Learning of Radiology Reports for Pulmonary Embolus: Is a Computer Reading My Report? , 2018, Radiology.

[67]  J. López-Goñi,et al.  Are previous suicide attempts a risk factor for completed suicide? , 2018, Psicothema.

[68]  Jinhee Lee,et al.  Development of a suicide index model in general adolescents using the South Korea 2012–2016 national representative survey data , 2019, Scientific Reports.