Improving operational forecasting effectiveness through seasonal time series models: a case study at the heart failure clinic, University Medical Center Ho Chi Minh City

Abstract

Background: The Heart Failure Clinic at University Medical Center Ho Chi Minh City faces rising patient demand and uneven session workloads, pressuring staffing, waiting times, and service quality. Objectives: This project evaluated seasonal time-series forecasting models and translated results into practical management solutions for the clinic. Methods: A mixed-methods design combined qualitative interviews with clinic managers and staff with quantitative analysis of 535 session-level observations (August 2022–December 2025). SARIMA and Holt-Winters models were compared using RMSE, MAE, and MAPE. Findings: Both models demonstrated short-term forecasting utility. SARIMA (0,1,2) (1,0,1)12 was selected for superior reliability and residual diagnostics. First-quarter 2026 forecasts projected an average of 20.76 patients per session with reduced variability, indicating a stable yet persistently high workload. Saturday sessions carried the greatest overload risk. Conclusion: Four management solutions were proposed: buffer-based capacity control with weekday-specific staffing; patient flow optimization and demand smoothing; fast-track pathways for stable heart failure patients; and adaptive workforce development. Seasonal forecasting models offer practical value for outpatient clinic management by improving resource allocation, reducing waiting times, and supporting evidence-based decision-making.