Ai-Driven Workforce Analytics For Employee Retention in Healthcare Organizations: Predicting Turnover Risk and Optimizing Strategic Human Resource Decisions

Abstract

Employee turnover presents a persistent challenge for healthcare organizations because workforce instability can affect operational continuity, staffing capacity, and human resource planning. This study examined the potential of AI-driven workforce analytics for predicting employee turnover and supporting strategic retention decisions. A healthcare workforce dataset comprising 1,676 employees was analyzed using demographic, occupational, compensation, satisfaction, and career-related characteristics. Employee attrition was treated as the prediction outcome, and Logistic Regression, Random Forest, and Gradient Boosting models were developed using a stratified training-test approach. Model performance was assessed using accuracy, precision, recall, F1-score, and ROC-AUC. Overall attrition was 11.87%, with employees experiencing turnover characterized by younger age, lower monthly income, shorter organizational tenure, lower job satisfaction, and poorer work-life balance. Overtime showed a marked association with turnover, with attrition rates of 29.2% among employees working overtime and 5.0% among those without overtime. Gradient Boosting achieved the highest accuracy (0.909) and ROC-AUC (0.911), while Logistic Regression achieved the highest recall (0.760). The findings demonstrate that predictive workforce analytics can identify meaningful turnover patterns while supporting risk-informed retention planning. Integrating predictive evidence with workforce characteristics may help healthcare organizations prioritize workload management, employee engagement, compensation, career development, and work-life balance initiatives.