An Optimization-based Decision Support Framework for Scheduling Problems in Healthcare: Addressing Outpatient Pathways and Workforce Management

This thesis proposes an optimization-based decision support framework to address key scheduling challenges in healthcare, focusing on outpatient care and workforce management with limited resources. Motivated by the needs of chronic patients with Non-Communicable Diseases (NCDs) and care managers requiring centralized coordination, the research introduces the novel NCDs Agenda problem: mid-term multi-appointment scheduling of care pathways, aiming to maximize delivered services while avoiding external referrals. A major contribution is the first application of Logic-Based Benders Decomposition in Answer Set Programming (ASP), with enhanced Benders cuts using unsatisfiable cores to efficiently solve large instances. To model daily operations, a new open-shop scheduling variant is proposed. Addressing duration uncertainty, a novel Robust Optimization approach is presented which manages delay propagation across both machine (operator) and job (patient) schedules, in case of multi-operation jobs. A state-based model incorporating Dynamic Programming is developed and validated through simulation. Finally, a real-world shift scheduling problem is tackled using ASP-based fairness mechanisms, targeting workload imbalances among highly specialized personnel, vital for staff satisfaction and sustainable management. Novel memory-based mechanisms are proposed to promote equity and validated on real data by staff members, showing improved service coverage and workload balance.