A Neuro-Surgery Ward Bed-Allocation Modeling System using Software Agents

R. Paranjape, M. Ogrady, and S.G. Nejad (Canada)


Modelling, simulation, artificial intelligence, and neural networks


This paper presents a software-agent system for modeling bed usage in a Neuro-Surgery ward at a regional Canadian hospital. The system is based on actual utilization data from the period March 1, 2005 – March 29, 2006. The system uses software agent technology to represent human patients in the Neuro-Surgery ward. A patient agent generator is used to produce a flow of patient agents for entry into the ward. Patient agents claim beds in the Neuro-Surgery ward (and if necessary in the overflow of the ward) of the Hospital. The patient agents remain in their claimed beds for the number of days that the patient agent is assigned, before moving out of the ward and out of the health care model. The length of stay (LOS) in the hospital bed is determined through the patient illness and the patient treatment as set by the assigned patient physician. The system is focused on creating a model that will allow an effective study of the dynamics of hospital bed allocation and usage in a specific ward. The system is an initial attempt to allow hospital administrators to examine ‘what if’ scenarios such as population increases, changed in physician treatment policy, etc. and the impact of these changes on resources and capacity of the hospital ward.

Important Links:

Go Back