Short-term physician rescheduling model with feature-driven demand for mental disorders outpatients

Physician rescheduling for a short term is common and critical when the guideline schedule for a long-term horizon needs adjustments. However, this topic is less studied in the literature. In this study, we develop a physician rescheduling model (PRM) with capacity allocation for outpatients. As in practice, the stochastic demand is assumed to be nonstationary and driven by auxiliary features. The capacity shortage risk is measured by the certainty equivalent that has shortage probability and expected shortage guarantees. To mitigate the capacity shortage risk, the Minimized Risk Tolerance Level (MRTL) decision criterion against the demand uncertainty is adopted for risk-averse hospital managers. Based on that, a bi-objective formulation that jointly considers PRM and MRTL is proposed to balance the operational costs and capacity shortage risk. We establish the tractability by an exact iterative algorithm that exploits the integer property of capacity and show that it finds all Pareto solutions. The case studies based on the real-life data of mental disorders demonstrate that our PRM performs well in handling the tradeoff between the capacity shortage risk and rescheduling costs.