考虑急诊随机需求的CT室预先排程研究

Advance Scheduling of CT Department under Emergency Random Demand

在考虑急诊随机插队情形下,以服务系统期望总收益最大化为目标建立马尔科夫决策模型描述CT室预先排程问题.精细划分预约病人的费用信息,精确刻画预约病人的直接等待时间.运用收益管理理论,通过边际分析技巧获得各等级病人的预约限制数量,保证接受预约请求的增量效益大于0.数值算例结果显示,得到的预约策略在医院服务病人的收入、病人的直接等待成本、预约请求被推迟的惩罚成本、医疗资源加班成本和空闲成本之间达到较好的平衡.由模型的结构性质知,得到的策略在实践中容易操作.

The advance scheduling problem of CT department under emergency random demand is described as a MDP in order to improve the expected total revenue of the service system.Emergency patients arrive randomly but should be served immediately.Regular patients are divided into several classes based on medical fees.The direct waiting time of regular patients is explicitly described.The booking limit number for each patient class is given by marginal analysis.The booking limit numbers ensure that the incremental benefit of accepting a request is positive.Numerical analysis shows that our advance scheduling policy could achieve balance between the income of hospital,the direct waiting cost for regular patients,penalty for delaying regular patients’ requests,overtime cost of medical resources and penalty for idling of medical resources.It is easy to put our capacity management policy into practice.