A pandemic is one of the types of natural disasters that have always caused great and sometimes irreparable damage to human societies throughout history. Since the statistics presented about the victims of epidemics show unbelievable figures, research and activity about dealing with this disaster and reducing its casualties is essential. Lack of timely treatment and prolonged patient waiting time is the most important factor that increases the mortality rate during disease outbreaks. The admission control problem is one of the problems that exist in the field of healthcare and can have a great effect on reducing patients' expectations. In this regard, in this research, the admission control problem to reduce the waiting time is discussed. Unlike previous studies that investigated this problem only at the level of one hospital, this research is trying to reduce patients' waiting time at the level of multi-hospitals. The main objective of this research is that during the disease outbreak, patients are admitted and hospitalized in the appropriate hospital to reduce congestion in hospitals. For this purpose, two Markov Decision Process (MDP) models are presented. In the first model, the number of beds is considered as a parameter, and in the second model, the number of inpatient beds is considered as a decision variable in the problem. The reason for presenting two models for the problem in question can be stated as follows: In some cases, hospital officials have to make decisions regarding how to allocate beds to the inpatient departments of pandemic and non-pandemic patients and determine the number of beds in each of these two departments. At other times, the bed allocation problem is not raised and the number of beds in these two departments is constant during outbreaks. By considering small instances of the problem, the proposed models solved exactly with a dynamic programming algorithm, and appropriate policies extracted about the allocation of pandemic patients to hospitals. The obtained policies eva‎luated with other heuristic policies by relatively large instances of the problem and with the help of simulation tools. The simulation results illustrated that the establishment of the admission control unit can have a great effect on improving the waiting time of patients and other performance measures. Moreover, if the item of equality is considered, the policies extracted from the dynamic programming method perform better than other heuristic policies, and if this item is ignored, some heuristic policies perform better.

فرشته پرورش

رضا زنجيراني فراهاني

مرتضي راستي برزكي , مهدي ايران پور