Case Mix Index weighted multi-objective optimization of inpatient bed allocation in general hospital

Inpatient bed, as one of the most critical resources for general hospitals, has to be effectively allocated among different departments to achieve the goal of preferable balance between patient service and resource utilization. To address this issue, studies in the past have been focusing on the development of bed allocation optimization algorithms with the objective of improved patient admission rate or higher bed occupancy rate. But in the context of hierarchical medical system promoted by the medical reform currently ongoing in China, the level of disease severity or treatment difficulty of clinical cases also has to be involved in the evaluation of the hospital performance, especially for the highly ranked national-level general hospitals. Case Mix Index (CMI), as an internationally recognized index which is highly correlated with the level of clinical complexity, is thus introduced to evaluate hospitals uniformly and facilitate straightforward comparison among them. To be aligned with the new requirement for an improved CMI, a multi-objective comprehensive learning particle swarm optimization algorithm is proposed in this paper based on an upgraded queueing model which incorporates CMI as an important weight into the traditional optimization objectives. Experimental results indicate that a higher CMI can be achieved with the new method and meanwhile both patient admission rate and bed occupancy rate with which hospital managers are still concerned will not be much influenced. The method is developed for a tertiary public hospital in Shanghai and can also be applied by hospital managers in other hospitals of the similar scale.