[Predictive Capacity of a Clinical Questionnaire to Identify the Propensity for Occupational Mental Health Disorders]

Area of Science:

• Occupational health
• Machine learning applications
• Psychological disorder assessment

Context:

• Chilean Law No. 16.744 regulates work-related mental health disorders.
• Current qualification processes face challenges in precision and objectivity.
• Need for enhanced diagnostic tools in occupational mental health.

Purpose:

• To evaluate the feasibility of machine learning (ML) algorithms for supporting mental health qualification.
• To assess the accuracy and efficiency of ML models in identifying work-related mental health conditions.

Summary:

• A mental health disease propensity questionnaire was administered to 340 Chilean workers.
• Decision tree and logistic regression models, including enhanced "Plus" versions, were evaluated.
• The "Plus Tree" model achieved 91.2% accuracy, with improved performance on balanced datasets.

Impact:

• ML algorithms can enhance the efficiency and objectivity of mental health disorder qualification.
• Further database expansion and model refinement are crucial for improving diagnostic precision.
• Potential for more accurate and reliable assessment of occupational mental health claims.