The growing complexity of healthcare management in the digital economy has heightened the need for intelligent, data-driven approaches to enhance operational efficiency and strategic decision-making. This study presents a machine learning (ML)-driven framework for optimising management efficiency and forecasting cost savings in the Big Health sector using simulation-based data. The research adopts a quantitative experimental methodology, implementing four advanced ML models, Random Forest, Support Vector Machine (SVM), XGBoost Regressor, and Linear Regression, developed and executed in Python (Google Colab). Simulated data representing healthcare operations were generated to address real-world constraints around data accessibility and privacy. The Random Forest classifier achieved a high accuracy of 99.67% and an F1 Score of 0.994 in identifying efficient management practices. For predictive analysis, XGBoost outperformed Linear Regression with an RMSE of 1173.41 and an MAE of 933.44 in forecasting future cost savings. The results confirm that ensemble ML models trained on realistic, simulation-based datasets can significantly support strategic decision-making in healthcare environments. This study recommends the use of ML-simulation pipelines as decision-support tools, offering practical implications for digital health management. However, the reliance on synthetic data presents limitations, calling for future validation with real-world datasets.