Dynamic Capacity Allocation for Integrated Online and Offline Outpatient Services With Stochastic Multiple Revisits

Internet healthcare provides a new access way for revisits through texts or videos, which can lighten the service load on offline hospitals. In this study, we investigate the integrated capacity allocation problem while incorporating multiple revisit transitions between online and offline outpatient systems. The heterogeneity of revisit patients in terms of their chronic conditions and disease progression is also thoroughly considered. We formulate the problem as a multistage stochastic mixed-integer programming (SMIP) model. Dynamic decisions of matching service capacities with stochastic demands for first visit and multitype multiple revisits are made at each stage, which ensure that each type of revisit is assigned appointment on the patient preferred day as much as possible under meeting the interval restrictions of two consecutive revisits. We reformulate the model into an equivalent formulation that can be solved directly and develop a decomposition-based adaptive capacity allocation with revisit priority algorithm to solve the model. Numerical results illustrate the superiority of our proposed approach in terms of computation time and solution quality compared with commercial solver. In addition, to aid practitioners in applying the decision-making model more effectively, we also provide managerial insights into key factors such as capacity and demand patterns, revisit intervals, and cost coefficients. For instance, managers can identify and implement the optimal capacity pattern for a given demand pattern, as well as the best combinations of demand and capacity patterns, to minimize operation costs.