Integrating simulation with simulated annealing for scheduling physicians in an understaffed emergency department

Abstract

In this paper, a physician scheduling problem in an emergency department (ED) is studied. A simulation-optimisation approach, which integrates simulation into simulated annealing is proposed to solve the problem. The algorithm explores solutions iteratively, where simulations are carried out to evaluate the solution quality at each step. With the proposed simulation-optimisation algorithm, a case study of an understaffed ED in Hong Kong is conducted. The historical data provided by the ED are used to draw some managerial insights into emergency-physician scheduling. It is found that (i) the best staffing level has a similar profile to the patient arrival rate, but shifts some period behind; (ii) the use of staggered shifts is helpful in reducing patient waiting time; (iii) physician scheduling decisions have a greater impact particularly when the ED is small; and (iv) ED pooling seems more beneficial than having a good physician schedule.