Estimating Business Loss to a Hospital Emergency Department from Patient Reneging by Queuing-Based Regression

摘要

Viewing a hospital emergency department (ED) as a production network, patients arrive to the processing queue and may renege (leave) before entering service. Reneging is not just a function of the delay in the queue or the number in the queue, but also the attitude of the patients toward perceived delay or fullness which vary between EDs. In this paper, we present evidence that reneging in an ED queue can be well estimated by nonlinear regression on p{sub}K, the 'full queue' formula of the M/M/1/K, over a wide range of ED situations. The regression curve is used for estimating business loss to production in the ED. We argue that the curve is useful outside the range of data used for fitting because the curve's functional form is not ad hoc as is most regression; instead it represents a universal queuing behavior.