Collaborative optimization of surgery and ICU under uncertainty

Abstract

Due to the COVID-19 pandemic, hospitals are facing an overload of staff and medical equipment. To uphold the interests of surgeons, patients, and hospitals, it is crucial to allocate medical resources effectively and rationally. This study optimizes the operating room (OR) and intensive care unit (ICU) modules collaboratively, dynamically integrating their mutual constraints. The optimization considers objectives related to both patients and hospitals while also respecting surgeons’ scheduling preferences and reducing their overtime. Compared with existing research that focuses on the intra-day surgery scheduling and highlight the uncertainty of the upstream (i.e., surgery duration), we emphasize the uncertainty of upstream and downstream (number of emergency patients, surgery duration and length of stay), and adopts a multi-day scheduling plan to provide practical operation scheduling for hospitals to allocate surgeons, ORs and ICU beds rationally. Due to the complexity of the model, an improved Non-dominated Sorting Genetic Algorithm-II based on fuzzy theory is proposed to solve the problem. We evaluate our method using a real case from Lyon Hospital in France and conduct an extensive sensitivity analysis on the parameters. Compared to the stochastic optimization model, our model reduces the OR costs and ICU bed by 8.33% and 4.93%, respectively. Additionally, it improves surgeon satisfaction by 38.54% and patient satisfaction by 31.90%.