Access the full text.
Sign up today, get DeepDyve free for 14 days.
K Hawton, H Bergen, D Casey (2007)
Self-harm in England: a tale of three cities. Multicentre study of self-harmSoc Psychiatry Psychiatr Epidemiol, 42
NJ Carson, B Mullin, MJ Sanchez (2019)
Identification of suicidal behavior among psychiatrically hospitalized adolescents using natural language processing and machine learning of electronic health recordsCurr Psychiatry Rep, 14
K Hawton, H Bergen, D Casey (2007)
Self-harm in England: a tale of three citiesCrisis, 42
JS Obeid, J Dahne, S Christensen (2020)
Identifying and predicting intentional self-harm in electronic health record clinical notes: deep learning approachSci Rep, 8
S Velupillai, H Suominen, M Liakata (2018)
Using clinical Natural Language Processing for health outcomes research: overview and actionable suggestions for future advancesJ Affect Disord, 88
J Robinson, K Witt, M Lamblin (2020)
Development of a self-harm monitoring system for VictoriaInt J Environ Res Public Health, 17
NJC Stapelberg, M Randall, J Sveticic, P Fugelli, H Dave, K. Turner (2021)
Data mining of hospital suicidal and self-harm presentation records using a tailored evolutionary algorithmNatl Health Stat Report, 3
J Robinson, K Witt, M Lamblin (2020)
Development of a self-harm monitoring system for VictoriaJ Artif Intell Res, 17
S Sperandei, A Page, MJ Spittal, K Witt, J Robinson, J. Pirkis (2020)
Using the ?presenting problem? field in emergency department data improves the enumeration of intentional self-harm in NSW hospital settings
S. Bachmann (2018)
Epidemiology of suicide and the psychiatric perspectiveInt J Environ Res Public Health, 15
SB Choi, W Lee, J-H Yoon, J-U Won, DW. Kim (2018)
Ten-year prediction of suicide death using Cox regression and machine learning in a nationwide retrospective cohort study in South KoreaPharmacoepidemiol Drug Saf, 231
S Hiles, H Bergen, K Hawton, T Lewin, I Whyte, G. Carter (2015)
General hospital-treated self-poisoning in England and Australia: comparison of presentation rates, clinical characteristics and aftercare based on sentinel unit dataJ Psychosom Res, 78
DL Zahl, K. Hawton (2004)
Repetition of deliberate self-harm and subsequent suicide risk: Long-term follow-up study of 11 583 patientsBr J Psychiatry, 185
H Hedegaard, M Schoenbaum, C Claassen, A Crosby, K Holland, S. Proescholdbell (2018)
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 dataJ Am Board Fam Med
K Hawton, D Zahl, R. Weatherall (2003)
Suicide following deliberate self-harm: Long-term follow-up of patients who present to a general hospitalBr J Psychiatry, 182
IJ Perry, P Corcoran, AP Fitzgerald, HS Keeley, U Reulbach, E. Arensman (2012)
The incidence and repetition of hospital-treated deliberate self harm: findings from the world's first National RegistryPLoS One, 7
GS Collins, JB Reitsma, DG Altman, KGM. Moons (2015)
Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): the TRIPOD statement, 350
AEW Johnson, TJ Pollard, L Shen (2016)
MIMIC-III, a freely accessible critical care database, 3
HD Anderson, WD Pace, E Brandt (2015)
Monitoring suicidal patients in primary care using electronic health records, 28
S Horng, DA Sontag, Y Halpern, Y Jernite, NI Shapiro, LA. Nathanson (2017)
Creating an automated trigger for sepsis clinical decision support at emergency department triage using machine learningJMIR Med Inform, 12
R Blagus, L. Lusa (2013)
SMOTE for high-dimensional class-imbalanced dataBMC Bioinformatics, 14
NV Chawla, KW Bowyer, LO Hall, WP. Kegelmeyer (2002)
SMOTE: synthetic minority over-sampling techniqueAust N Z J Psychiatry, 16
R Blagus, L. Lusa (2013)
SMOTE for high-dimensional class-imbalanced dataMach Learn Appl, 14
K Witt, J. Robinson (2019)
Sentinel surveillance for self-harm: existing challenges and opportunities for the future, 40
A Müller, L Claes, D Smits, E Brähler, M. de Zwaan (2016)
Prevalence and correlates of self-harm in the German general populationSoc Psychiatry Psychiatric Epidemiol, 11
AC Fernandes, R Dutta, S Velupillai, J Sanyal, R Stewart, D. Chandran (2018)
Identifying suicide ideation and suicidal attempts in a psychiatric clinical research database using natural language processingPLoS One, 8
S Graham, C Depp, EE Lee (2019)
Artificial intelligence for mental health and mental illnesses: an overviewJ Biomed Inform, 21
JT Walkup, L Townsend, S Crystal, M. Olfson (2012)
A systematic review of validated methods for identifying suicide or suicidal ideation using administrative or claims dataPLoS One, 21 Suppl 1
ObjectiveAccurate identification of self-harm presentations to Emergency Departments (ED) can lead to more timely mental health support, aid in understanding the burden of suicidal intent in a population, and support impact evaluation of public health initiatives related to suicide prevention. Given lack of manual self-harm reporting in ED, we aim to develop an automated system for the detection of self-harm presentations directly from ED triage notes.Materials and methodsWe frame this as supervised classification using natural language processing (NLP), utilizing a large data set of 477 627 free-text triage notes from ED presentations in 2012–2018 to The Royal Melbourne Hospital, Australia. The data were highly imbalanced, with only 1.4% of triage notes relating to self-harm. We explored various preprocessing techniques, including spelling correction, negation detection, bigram replacement, and clinical concept recognition, and several machine learning methods.ResultsOur results show that machine learning methods dramatically outperform keyword-based methods. We achieved the best results with a calibrated Gradient Boosting model, showing 90% Precision and 90% Recall (PR-AUC 0.87) on blind test data. Prospective validation of the model achieves similar results (88% Precision; 89% Recall).DiscussionED notes are noisy texts, and simple token-based models work best. Negation detection and concept recognition did not change the results while bigram replacement significantly impaired model performance.ConclusionThis first NLP-based classifier for self-harm in ED notes has practical value for identifying patients who would benefit from mental health follow-up in ED, and for supporting surveillance of self-harm and suicide prevention efforts in the population.
Journal of the American Medical Informatics Association – Oxford University Press
Published: Dec 13, 2021
Keywords: self-harm; natural language processing; machine learning; emergency department; suicidal ideation
Read and print from thousands of top scholarly journals.
Already have an account? Log in
Bookmark this article. You can see your Bookmarks on your DeepDyve Library.
To save an article, log in first, or sign up for a DeepDyve account if you don’t already have one.
Copy and paste the desired citation format or use the link below to download a file formatted for EndNote
Access the full text.
Sign up today, get DeepDyve free for 14 days.
All DeepDyve websites use cookies to improve your online experience. They were placed on your computer when you launched this website. You can change your cookie settings through your browser.