Ten-year prediction of suicide death using Cox regression and machine learning in a nationwide retrospective cohort study in South Korea

Seung-Hun, Choi; Lee, Wanhyung; Yoon, Jin Ha; Won, Jong Uk; Kim, Deok Won

doi:10.1016/j.jad.2018.01.019

Cited by 65 publications

(63 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Oh et al [ 22 ] classified individuals with a history of suicide attempts among patients with depression or anxiety disorders by applying artificial neural networks to multiple psychiatric scales and sociodemographic data with an accuracy of 87–91%. Moreover, a recent study investigated the probability of death by suicide using general characteristics and insurance data from the National Health Insurance Service cohort in Korea, showing fair performance (AUC=0.68) of machine learning models in predicting death by suicide [ 23 ]. In the present study, we intended to develop a machine learning model predicting suicide risk in the general population.…”

Section: Discussionmentioning

confidence: 99%

Use of a Machine Learning Algorithm to Predict Individuals with Suicide Ideation in the General Population

Ryu¹,

Lee²,

Lee³

et al. 2018

Psychiatry Investig

View full text Add to dashboard Cite

ObjectiveIn this study, we aimed to develop a model predicting individuals with suicide ideation within a general population using a machine learning algorithm. MethodsAmong 35,116 individuals aged over 19 years from the Korea National Health & Nutrition Examination Survey, we selected 11,628 individuals via random down-sampling. This included 5,814 suicide ideators and the same number of non-suicide ideators. We randomly assigned the subjects to a training set (n=10,466) and a test set (n=1,162). In the training set, a random forest model was trained with 15 features selected with recursive feature elimination via 10-fold cross validation. Subsequently, the fitted model was used to predict suicide ideators in the test set and among the total of 35,116 subjects. All analyses were conducted in R. ResultsThe prediction model achieved a good performance [area under receiver operating characteristic curve (AUC)=0.85] in the test set and predicted suicide ideators among the total samples with an accuracy of 0.821, sensitivity of 0.836, and specificity of 0.807. ConclusionThis study shows the possibility that a machine learning approach can enable screening for suicide risk in the general population. Further work is warranted to increase the accuracy of prediction.

show abstract

Section: Discussionmentioning

confidence: 99%

Use of a Machine Learning Algorithm to Predict Individuals with Suicide Ideation in the General Population

Ryu¹,

Lee²,

Lee³

et al. 2018

Psychiatry Investig

View full text Add to dashboard Cite

show abstract

“…On the other hands, one previous study based on a Korean national health insurance database reported that patients who visited a medical facility for injury, poisoning, and certain other consequences of external causes (S00-T98) had a significantly increased risk of suicide death. 29 This difference could be explained that severity of injury between primary care and emergency center. In the earlier study, disease information was solicited from all types of medical facilities, including psychiatric departments and emergency departments.…”

Section: Discussionmentioning

confidence: 99%

Association between Suicidal Ideation and Causes of Primary Care Visits: A Cross-Sectional Study to Identify Patients with Suspected Suicidal Ideation

Lee

Kim

2020

Psychiatry Investig

View full text Add to dashboard Cite

Objective Suicide victims are likely to visit a primary care provider within 1 month of the suicide attempt. However, it is impossible for primary care providers to screen all patients for suicidal ideation or attempts. We, therefore, aimed to investigate potential differences in the chief diagnoses received by primary care patients with and without suicidal ideation, as this information may provide clues and predict patients at risk of suicide.Methods This cross-sectional study included 1,211 Korean adults aged ≥20 years who had participated in the Korea National Health and Nutrition Examination Survey during 2014 and provided information about their histories of suicidal ideation and visits to primary care providers. Multiple logistic regression analyses were used to examine whether suicidal ideation was associated with specific diseases after controlling for age, sex, household income, education, occupation, marital status, smoking, alcohol consumption, and physical activity.Results Among primary care patients, the sample-weighted percentage of suicidal ideation was approximately 12%. The diagnoses of neoplasm and mental disorder were associated with a higher likelihood of suicidal ideation {odds ratio (OR) [95% confidence interval (CI)] 8.32 (1.01–68.56) and 6.19 (1.72–22.33), respectively}. Notably, diseases of the musculoskeletal system and connective tissue were also associated with suicidal ideation [OR (95% CI): 1.93 (1.09–3.43)].Conclusion The observed visit patterns and correlations of certain diagnoses with suicidal ideation suggest that primary care physicians should pay attention to patients with relevant disorders, consider screening suicidal ideation, and properly consult to a psychiatrist.

show abstract

“…Prior suicide attempts were identified from either prior admission with the codes for suicide=self-harm or the ICD-10 codes for comorbidities in suicide cases, following previous studies. 17,18 Clinical practices, psychiatric interventions, and hospital characteristics were also identified. Psychiatric interventions during hospitalization included psychiatric ward admission, psychotherapy with a psychiatrist, consultation-liaison psychiatry, and the use of antidepressants.…”

Section: Variablesmentioning

confidence: 99%

Clinical Trajectories of Suicide Attempts and Self-harm in Patients Admitted to Acute-care Hospitals in Japan: A Nationwide Inpatient Database Study

Ohbe

Goto

Yamazaki

et al. 2021

Journal of Epidemiology

View full text Add to dashboard Cite

Background: For patients with suicide attempts or self-harm, acute-care hospitals often function as the primary or sole point of contact with the healthcare system. However, little is known about patient characteristics or clinical trajectories of suicide attempts and self-harm episodes among those admitted to acute-care hospitals. This study aimed to describe the characteristics of suicide attempts and self-harm among patients admitted to acute-care hospitals, and the clinical practices provided in these hospitals, using a nationwide inpatient database in Japan. Methods: Using data from the Japanese Diagnosis Procedure Combination inpatient database from June 2015 to March 2017, we identified patients with emergency admission for suicide attempts or self-harm. We did not include patients with elective admission to psychiatric hospitals or outpatients. We described patient characteristics, treatments for physical injuries, psychiatric interventions, and discharge status. Results: We identified 17,881 eligible patients during the 22-month study period. Overall, 38% of the patients did not have any psychiatric or behavioral comorbidities at admission. The most common suicide method was drug overdose (50%), followed by hanging (18%), jumping from a height (13%), cutting or piercing without wrist cutting (7.1%), poisoning (6.6%), and wrist cutting (5.4%). Suicide was completed by 2,639 (15%) patients. Among patients discharged to home, 51% did not receive any psychiatric intervention. In 468 acute-care hospitals (54%), no psychiatric intervention was provided during the study period. Conclusion: We found that half of acute-care hospitals did not provide any hospital-based psychiatric care for patients with suicide attempts or self-harm.

show abstract

Ten-year prediction of suicide death using Cox regression and machine learning in a nationwide retrospective cohort study in South Korea

Cited by 65 publications

References 29 publications

Use of a Machine Learning Algorithm to Predict Individuals with Suicide Ideation in the General Population

Use of a Machine Learning Algorithm to Predict Individuals with Suicide Ideation in the General Population

Association between Suicidal Ideation and Causes of Primary Care Visits: A Cross-Sectional Study to Identify Patients with Suspected Suicidal Ideation

Clinical Trajectories of Suicide Attempts and Self-harm in Patients Admitted to Acute-care Hospitals in Japan: A Nationwide Inpatient Database Study

Contact Info

Product

Resources

About