Essays on behavioral operations in the emergency departments

Bebitoglu, Basak (2023) Essays on behavioral operations in the emergency departments. Doctoral thesis, University of London: London Business School.

Abstract

Understanding how servers make operational decisions in labour intensive service settings is important for managing operations optimally as well as providing a high-quality service. This thesis presents three studies that explores behavioural operations within the context of an Emergency Department and its implications for practice. Extant literature assumes service discipline is an engineering decision that is set upfront to achieve a certain objective.

The first study questions this assumption and examine practitioners' decisions in setting patient priorities. Our results show that service providers exercise discretion over the service discipline to balance the trade-off between fairness and efficiency. On the one hand, they strive to be fair by prioritizing patients that waited the longest when the system is not congested. On the other hand, they try to maintain efficiency by deviating to short service time patients based on system's condition.

Inspired by these results, in the second study, we develop a combination of analytic and simulation models to determine the impact of congestion-based service discipline on system performance. We analyse a system that follows first-come-first-served service discipline while the number of people in the system is below a certain threshold and prioritizes short service times above it. Our results show that one can reduce average delays by reducing the threshold. However, this happens at the expense of increasing delays for customers with long service times.

In the third study, we look at the impact of top-down waiting time targets on patient length-of-stay. The UK government mandates 95% of the patients to be seen, admitted or discharged within four hours of their arrival. We find, by using a dynamic model, that this causes a speed up of the treatment as patients get closer to the target. This implies that having such objective strains resources and may decrease patient-doctor interaction time, which in return might also have health consequences.