Predicting the Risk of Burnout Syndrome Using Korean Occupational Stress Scale (KOSS): A Machine Learning Approach.

Area of Science:

• Occupational health
• Psychiatry
• Data science

Background:

• Workplace changes increase occupational stress, leading to burnout syndrome (BOS).
• The Korean Burnout Syndrome Scale (KBOSS) faces limitations due to stigma and low compliance.
• A need exists for objective, scalable methods to assess BOS risk.

Purpose of the Study:

• Develop machine learning (ML) models to predict BOS risk.
• Identify key occupational stress factors contributing to BOS.
• Provide a scalable alternative to traditional BOS assessment tools.

Main Methods:

• Evaluated five ML algorithms on a dataset of 1,205 individuals from 40 companies.
• Utilized resampling and 5-fold grid search cross-validation for model optimization.
• Employed SHAP analysis to identify influential occupational stress factors and quantify their contribution.

Main Results:

• All five ML models showed strong predictive performance, with Random Forest achieving a ROC-AUC of 0.904.
• SHAP analysis identified "Job instability" and "Lack of reward" as primary BOS risk factors.
• Critical transition points in KOSS factor responses were identified, correlating with BOS risk.

Conclusions:

• ML models can effectively predict BOS risk from occupational stress factors.
• Early identification of at-risk employees enhances cost efficiency and intervention scalability.
• This ML-driven approach offers a promising solution for proactive BOS management.