Identification of Suicide Attempt Risk Factors in a National US Survey Using Machine Learning

Area of Science:

• Psychiatry and Behavioral Sciences
• Public Health
• Data Science and Machine Learning

Background:

• Over one-third of individuals who attempt suicide do not receive mental health treatment.
• Identifying risk factors in the general population is crucial for effective suicide prevention strategies.

Purpose of the Study:

• To identify suicide attempt risk factors in the general US adult population.
• To utilize a data-driven machine learning approach with a comprehensive survey dataset.

Main Methods:

• Analysis of data from the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) waves 1 and 2.
• Development of a suicide attempt risk model using a balanced random forest algorithm with over 2500 survey questions.
• Assessment of model performance using cross-validation, area under the receiver operator characteristic curve, sensitivity, and specificity.

Main Results:

• The machine learning model achieved an area under the curve of 0.857, with 85.3% sensitivity and 73.3% specificity.
• Key risk factors identified include previous suicidal ideation/behavior, functional impairment from emotional problems, younger age, lower education, and recent financial crisis.
• The model identified 1.8% of the US population at a 10% or greater risk of suicide attempt.

Conclusions:

• Confirmed known risk factors like prior suicidal behaviors and ideation.
• Identified novel risk factors, including functional impairment due to mental disorders and socioeconomic disadvantage.
• Findings can inform clinical assessments and the development of new suicide risk scales.