Distributed patient scheduling in hospitals

Abstract

Patient scheduling in hospitals is a highly complex task. Hospitals have a distributed organisational structure; being divided into several autonomous wards and ancillary units. Moreover, the treatment process is dynamic (information about the patients’ diseases often varies during treatments, causing changes in the treatment process). Current approaches are insufficient because they either focus only on the single ancillary units, and therefore do not consider the entire treatment process of the patients, or they do not account for the distribution and dynamics of the patient scheduling problem. Therefore, we propose an agent based approach in which the patients and hospital resources are modelled as autonomous agents with their own goals, reflecting the decentralised structures in hospitals. In this multi-agent system, the patient agents compete over the scarce hospital resources. Moreover to improve the overall solution, the agents then negotiate with one another. To this end, a market mechanism is described, in which each self interested agent tries to improve its own situation. In particular we focus on how the agents can calculate demand and supply prices based upon their current schedule. Further, an evaluation of first results of the proposed method is given.

More information

Identifiers

Catalogue record

Export record