Identifying the suicidal ideation risk group among older adults in rural areas: Developing a predictive model using...

Area of Science:

• Geriatric mental health
• Machine learning applications in healthcare
• Public health surveillance

Background:

• Identifying suicidal ideation among older adults in rural areas presents unique challenges.
• Early detection of suicide risk is crucial for effective intervention and prevention strategies.
• Machine learning offers potential for developing sophisticated predictive models in healthcare.

Purpose of the Study:

• To develop and evaluate machine learning models for predicting suicidal ideation risk in rural older adults.
• To identify significant demographic and health factors associated with suicidal ideation in this population.
• To provide a tool for healthcare providers to facilitate early identification and intervention for suicide prevention.

Main Methods:

• An exploratory, descriptive, cross-sectional study design was employed.
• Data were collected from 650 older adults (aged over 65) in rural South Korea using self-report questionnaires.
• Machine learning methods, including decision tree, random forest, and logistic regression, were utilized for data analysis.

Main Results:

• Depression, pain, age, and loneliness were identified as significant predictors of suicidal ideation.
• The predictive models demonstrated good performance, with notable area under the receiver operating characteristic curve values.
• Model evaluations indicated moderate to high sensitivity and specificity in identifying at-risk individuals.

Conclusions:

• Machine learning models show promise for predicting suicidal ideation risk in older adults residing in rural settings.
• Incorporating screening for depression, pain, age, and loneliness is recommended for assessing suicide risk.
• These models can assist healthcare providers in clinical and community settings to initiate timely interventions, potentially preventing geriatric suicide.