Conectar
To combine the Job Demand-Resource (JD-R) model with machine learning (ML) techniques to identify the key factors affecting job burnout (JB) among Chinese nurses.
A Cross-Sectional Study.
This study utilised a stratified sampling method to recruit 3449 eligible nurses from eight cities in Shandong Province between June and December 2021. After data cleaning, 2998 valid samples were retained. The dataset was randomly split into a training set (75%) and a test set (25%). The Boruta algorithm was used to select relevant variables for model construction. Six-millilitre models were compared using cross-validation, with mean absolute error (MAE), root mean square error (RMSE) and R-squared (R 2) used to select the best model. The Shapley Additive Explanation (SHAP) method was used to identify key predictors of JB.
The average JB score among nurses was (32.88 ± 11.45). Among the 20 variables, 17 were identified by the Boruta algorithm as strongly associated with JB, including 7 job demand-related variables and 10 job resource-related variables. After comparing 6-ml models, the Random Forest was identified as the optimal model (MAE = 6.56, RMSE = 8.86, R 2 = 0.63). SHAP analysis further revealed the importance ranking of these 17 variables and identified four key predictors: psychological distress (SHAP = 4.07), perceived organisational support (SHAP = 2.03), emotional intelligence (SHAP = 1.81) and D-type personality (SHAP = 1.73).
By integrating the JD-R model framework, ML algorithms proved effective in identifying critical predictors of nurses' JB. SHAP analysis identified four primary determinants: psychological distress, perceived organisational support, emotional intelligence and D-type personality. These findings provide novel insights for nursing administrators to optimise intervention strategies.
Not applicable.
This study did not include patient or public involvement in its design, conduct or reporting.
FreshRSS