Abstract Submission

Hybrid heuristic for the patient-bed allocation problem

Invited abstract in session MB-31: Capacity planning in healthcare, stream ORAHS: OR in Health and Healthcare.

Area: OR in Health, Life Sciences and Sports

Monday, 10:30-12:00
Room: G108

Authors (first author is the speaker)

1. Fabian Schäfer
Supply and Value Chain Management, Technical University of Munich
2. Alexander Hübner
Supply and Value Chain Management, Technical University Munich
3. Dominik G Grimm
Bioinformatics, TUM Campus Straubing for Biotechnology and Sustainability, Weihenstephan-Triesdorf University of Applied Sciences


Managing patient to bed allocations is an everyday task in hospitals. In recent years it has moved into focus due to a general rise in occupancy levels and the resulting need to efficiently manage tight hospital bed-capacities. This holds true especially for large maximum-care hospitals, which are by definition obliged to treat any incoming patient. Hence, maximum-care hospitals exhibit a high ratio of emergency patients as well as a high volatility and uncertainty regarding patient arrivals and lengths of stay. The patient-bed allocation problem (PBA) decision support model refined the patient admission scheduling problem (PAS) by means of a real-world situation in a large maximum-care hospital. The PBA identifies three main stakeholders, namely patients, nursing staff, and doctors, whose individual objectives and constraints lead to a potential trade-off situation. Due to the combinatorial complexity of the PBA, there is a need for a heuristic that intelligently assists the bed manager in taking fast decisions and is able to deal with uncertain situations through quick recalculations. Therefore, we developed a hybrid heuristic based on a preferred iterative look ahead technique and a genetic algorithm. Furthermore, to deal with the high volatility and uncertainty of emergency admissions, we trained a deep neural network to forecast emergency occupations based on features related to time designation, weather, fairs, and holidays.


Status: accepted

Back to the list of papers