基于随机优化的门诊患者调度研究

After about three decades' healthcare reform, medical system of China has been greatly improved. However, the traditional problem of "the difficulty and high cost of getting medical treatment" has not been resolved now. To deal with the "difficulty" problem, this project aims at developing new models and methodologies for the outpatient scheduling. According to the time horizon, outpatient scheduling is divided into advanced scheduling (in terms of weeks or days) and appointment scheduling (in terms of hours or minutes). The features of different medical services, i.e., single, multi-reentrant, and combined diagnostic or treatments, are considered. Based on Markov decision process, stochastic programming, and other stochastic optimization methods, this project will deal with the following two issues: 1) advanced scheduling by considering the waiting time targets of outpatients, future patients' arrival, and uncertainties of combined medical resources; 2) appointment scheduling by considering uncertainties from both patients and physicians, different service time of new and reentrant patients, and combination of different medical services. Finally, case study will be performed by using the data from Ruijin Hospital, First Hospital of Shanghai, and other hospitals from Wuxi and Changzhou. This project will enrich the patient scheduling, which is important for both theory and application.

经过近三十年的医疗改革，中国医疗系统取得了一定的成就，但传统的"看病难、看病贵"问题一直未能得到有效解决。本项目针对"看病难"问题突出的门诊服务，从以下两个时间维度研究患者调度：以日或周为单位的提前调度和以小时或分钟为单位的预约调度，以减少患者从进行预约到接受医疗服务之间的等待时间。本项目基于随机规划、马尔科夫决策支持等随机优化方法，针对不同门诊服务的特点，即单次、重入或组合的医疗服务，考虑患者的可忍耐等待时间，未来可能到达的患者信息，组合服务的不确定性，进行患者提前调度研究；考虑医患双方的不确定因素，首诊、复诊患者不同的服务时间，患者所需的组合医疗服务，进行患者预约调度研究；最后结合上海瑞金医院、上海市第一人民医院，无锡、常州等多地医院的实际情况，进行多样化的案例研究。本项目将丰富完善医疗服务中的患者调度研究，具有重要的理论和应用价值。

经过近三十年的医疗改革，中国医疗系统取得了一定的成就，但传统的“看病难、看病贵”问题一直未能得到有效解决。本项目针对“看病难”问题突出的门诊服务，从以下两个时间维度研究患者调度：以日或周为单位的提前调度和以小时或分钟为单位的预约调度，以减少患者从进行预约到接受医疗服务之间的等待时间。本项目首先针对单次的医疗服务，包括门诊或检查，基于马氏决策过程和随机规划等方法，考虑了患者不同的等待时间要求，探讨了关键医疗资源的配置决策优化与患者的提前调度与预约调度；针对重入的医疗服务，如放疗或住院，将时间槽作为服务器，考虑患者不同的等待时间要求，进行了医疗资源配置研究，并进一步考虑医疗服务的公平性，考虑了可及公平性与响应公平性，建立了多目标随机规划模型，并探讨了求解算法；针对组合医疗服务或检查，基于马氏决策过程，研究了如何将患者安排至不同的检查时间（以天为单位），进一步研究了安排至同一天的多个检查，具体安排至哪个时间段，以减少患者的等待时间。最后结合上海瑞金医院、仁济医院等多医院的实际情况，进行多样化的案例研究。本项目丰富完善了医疗服务中的患者调度研究，具有重要的理论和应用价值。