Abstract
Outpatients in Japan are generally divided into two groups: scheduled patients and walk-ins. Each scheduled patient arrives at a hospital with an appointment. On the other hand, walk-ins arrive at a hospital unexpectedly. In general, the ratio of walk-ins to all outpatients is considerable. This paper focuses on the average waiting times of both scheduled patients and walk-ins under a fixed set of scheduled patients, and deals with the consultation sequencing problem. Even though the average waiting time for all patients is the same under a set of candidate sequences, the average waiting time of scheduled patients (and also that of walk-ins) usually depends on the sequence. Therefore, to realize the target average waiting times, there is a requirement to appropriately allocate the total waiting time of all patients between the two groups. The relationship between the waiting times of these two groups is approximated by a linear equation of which the coefficient is given by the average number of patients and the mean consultation time of each group. This paper also tries to visualize the planned consultation sequence to alleviate the stress of waiting. The planned sequence may include one or more reserved slots for scheduled but not yet arrived patients, and open slots for walk-ins. Each waiting patient can grasp the minimum and maximum number of patients in front of him or her in the sequence. When a scheduled patient arrives he or she is placed at one of the reserved slots. When a walk-in arrives, an open slot for the patient is selected randomly. The types of slots, however, are invisible to the patients. To evaluate the proposed method, four alternative selection rules are prepared. Each rule selects a patient dynamically when a physician becomes ready for the next consultation. The comparison of the proposed method with alternative dynamic selection rules via simulation experiments indicated that the performance of the proposed method is similar to the alternative rules in terms of the average waiting times.
