Towards resilient healthcare : an application to the large-scale outpatient system

Abstract

A resilient healthcare system can absorb internal variability within the system, adapt to external variability from the environments by adopting recovery strategies, and maintain a required key functionality. However, prior studies neglect its application to the large-scale outpatient system (LSOS) with internal and external variabilities. Internal variability includes complex patient flows and stochastic service time. External variability considers urgent demand surge and unexpected supply loss, e.g., physician absence, which can impede LSOS from providing sufficient service capacity. Existing literature has two major gaps. First, a well-developed model considering multi-stage services and stochastic service time in multiple departments is desired for LSOS. Second, prior studies mainly focused on scenario-based methods, which are hard to be generalized in different healthcare systems. To bridge these gaps and develop a resilient large-scale outpatient system, we presented four studies. First, a multi-stage outpatient network model with extensive patient routings and stochastic service time is built via discrete-event simulation (DES). The effects of two sources of internal variability on clinical efficiency are examined. Second, critical department analysis is performed for LSOS with external variability, precisely, supply enhancement, demand surge, and supply loss. Third, the maximal integrated service capacity (MISC) is derived as the key functionality of LSOS. Fourth, a scenario-free resilience assessment framework is developed upon the maximal imbalance and the optimal recovery solution. Besides, three recovery strategies are compared, namely, demand contraction, supply reconfiguration, and integrated solution, to find the optimal solution. The major results of each study are listed as follows. 1) Patient flow variability reduces patient satisfaction and increases resource utilization, while service time variability has the opposite effect. 2) Supply enhancement has a marginal benefit in bottlenecks but deteriorates patient satisfaction due to incoordination among departments. External variability reduces patient satisfaction and increases the number of spillover patients. Critical department rankings rely on time sessions. 3) MISC relies on the spillover tolerance level. The departmental capacity decline can benefit the LSOS’s integrated capacity. 4) With a minor reassignment of physicians, integrated solution performs the best among the three recovery strategies. The contributions of this thesis lie in the following two aspects. Scientifically, the understanding of the internal and external variability for large-scale outpatient systems is enhanced. We synthesized the definition of resilient healthcare systems and put forward a scenario-free measurement. Methodologically, a workable DES model considering complex variability is developed for LSOS. A discrete optimization via simulation (DOvS) model is generated and solved with the mononu-SA algorithm. A multi-stage model is built for the resilience assessment framework and solved by a three-layer algorithm.