An integrated approach for scheduling health care activities in a hospital

To effectively utilise hospital beds, operating rooms (OR) and other treatment spaces, it is necessary to precisely plan patient admissions and treatments in advance. As patient treatment and recovery times are unequal and uncertain, this is not easy. In response a sophisticated flexible job-shop scheduling (FJSS) model is introduced, whereby patients, beds, hospital wards and health care activities are respectively treated as jobs, single machines, parallel machines and operations. Our approach is novel because an entire hospital is describable and schedulable in one integrated approach. The scheduling model can be used to recompute timings after deviations, delays, postponements and cancellations. It also includes advanced conditions such as activity and machine setup times, transfer times between activities, blocking limitations and no wait conditions, timing and occupancy restrictions, buffering for robustness, fixed activities and sequences, release times and strict deadlines. To solve the FJSS problem, constructive algorithms and hybrid meta-heuristics have been developed. Our numerical testing shows that the proposed solution techniques are capable of solving problems of real world size. This outcome further highlights the value of the scheduling model and its potential for integration into actual hospital information systems.

为有效利用病床、手术室（OR）及其他诊疗空间，需提前对患者收治与诊疗方案进行精准规划。由于患者的治疗与康复时长存在不均性与不确定性，此类规划颇具难度。为此，本文提出一种先进的柔性作业车间调度（Flexible Job-shop Scheduling, FJSS）模型，将患者、病床、病房以及医疗活动分别视为作业、单机设备、并行设备与工序。本研究的创新之处在于，可通过单一集成化方案实现对整个医院的建模与调度。该调度模型能够在出现偏差、延误、推迟或取消情况时，重新计算各环节的时间安排。此外，模型还涵盖了多项高级约束条件，包括设备与活动的准备时长、活动间的转运时长、阻塞限制与无等待要求、时序与占用约束、用于提升鲁棒性的缓冲机制、固定活动与工序序列、任务释放时间以及严格的截止期限。为求解该FJSS问题，本文开发了构造式算法与混合元启发式算法。数值实验结果表明，所提出的求解方法可处理实际规模的调度问题，这进一步凸显了该调度模型的应用价值，以及其集成至实际医院信息系统的潜力。