Hospital Bedsrevenue Management With Multiple Classes Of Patients And Length‐of‐stay

Considering the unmatch between supply and demand as well as the inefficiency in medical resource allocation, this paper focused on hospital beds allocation to discuss a non‐pricing network revenue management problem with multiple classes of customers whose characters and behaviors contrasted with each other.More specifically, this paper assumed that the hospital serving patients within a time horizon had a definite number of beds. For each time period, there were three classes of patients arriving at the hospital and requiring for length‐of‐stay. Since revenues were collected by accommodating patients while costs were suffered on waiting patients and unserved lost patients, the hospital had to carefully balance between different classes of patients in different time periods in order to achieve maximum profit.This paper first constructed a basic model with strong assumptions for the beds allocation problem, analyzing its structures and properties. Regarding to different situations of demand, the paper then discussed deterministic demand case and stochastic demand case respectively, centering on problem formulation and algorithm development. Numerical examples were later given to show the efficiency of the algorithms.This paper also made an effort to extend the basic model to some more general cases by relaxing its assumptions. It is hoped that this paper could provide some useful insights to other resources allocation problems confronted in health care, airline booking, hotel reservation, etc.