Queue Management Case Study

Most hospitals and clinics address their patient queue using FIFO, LIFO or static priority queues. Very few address dynamic priority queues. As a result, patients are faced with long eating period, decreasing the whole quality of service of the hospital. In this paper, we present a different queue management system where-in the queue is automatically varied based on the severity of the patient. Our algorithm is very simple and straight forward as you will see in the third section of this paper. This priority itself is variable, based on the different health factors and hospital or clinic condition. We show that this strategy is beneficial in most of the situations and is quite effective in preventing starvation. INTRODUCTION

Healthcare

Patients arrive at the hospital and register themselves. Then they wait for the initial consultation with a doctor or a nurse who determines their severity and the Length of Stay. If the patient is detected to be fine, he or she is discharged immediately so as to avoid unnecessary crowd in the queue. If further attention of the doctor is required, the patient is added to the queue with the estimated LOS, severity of the disease and their arrival time at the hospital or the clinic. The arrival time is converted to 24 Hours format and then converted completely to minutes for easier calculation. For example, 8:32 PM will be converted first to 20:32 and then to 1232 (i.e., 20×60 + 32). Now the priority of the patient is calculated using the following

This system can be completely automated expect for the initial examinations done b y the doctor or the nurse. As artificial intelligence and it’s enabling technology get better, this may also be automated. The more data is feed to the system, the more accurately it will calculate the coefficients for Severity, Time of Arrival and Length of Stay. These coefficients are the key element of this model and error in their calculation will render the whole system useless. We found that the proposed strategies result in shorter average LOS compared to using FIFO, LIFO or the Static Priority method and they are able to scale well over peak days and peak hours. As a part of the future work, it is possible to implement the proposed model with different strategies such as Shortest Consultation Time First [1], Longest Time First, Round Robin, etc. The proposed model can also be applied in different fields such as the passport issue center. In passport scenario, the priority can be decided by measuring the time by which the person need the passport. This will be a fixed quantity and hence the computation load on the system will be