Factors Associated With Suicidal Ideation Among Persons With Disabilities in South Korea: Retrospective Observational...

Area of Science:

• Public Health
• Mental Health Research
• Machine Learning in Healthcare

Background:

• South Korea faces a high suicide rate, notably among individuals with disabilities.
• Suicidal ideation and suicide attempts show a strong correlation in research.
• Persons with disabilities experience disproportionately high rates of suicidal ideation.

Purpose of the Study:

• To investigate factors contributing to suicidal ideation among persons with disabilities in South Korea.
• To utilize machine learning models for predicting suicidal ideation.
• To identify key predictors for targeted mental health interventions.

Main Methods:

• Analysis of data from the 2020 National Survey on Persons with Disabilities (n=6832).
• Feature selection using random forest to identify top 100 predictors from 1394 variables.
• Training and evaluation of five machine learning models: logistic regression, SVM, random forest, XGBoost, and neural network.

Main Results:

• 12.1% of persons with disabilities reported suicidal thoughts in the past year.
• Significant predictors included intense sadness, disability-related difficulties, and health satisfaction.
• The random forest model achieved the highest predictive performance (AUC 0.905).

Conclusions:

• Identified critical predictors of suicidal ideation in persons with disabilities.
• Highlights the need for focused mental health interventions and suicide prevention strategies.
• Machine learning effectively identified key factors, with further research needed on causality.