EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

711. Predicting next week’s bed census: combining medical expertise with data.

Invited abstract in session WD-10: Admission and discharge, stream OR in Health Services (ORAHS).

Wednesday, 14:30-16:00
Room: 11 (building: 116)

Authors (first author is the speaker)

1. Hayo Bos
University of Twente
2. Stef Baas
University of Twente
3. Richard Boucherie
Stochastich Operations Research, University of Twente
4. Erwin W. Hans
Industrial Engineering & Business Information Systems, University of Twente, fac. Behavioral Management and Social Science
5. Gréanne Leeftink
CHOIR, University of Twente

Abstract

Bed census predictions play a key role in hospital capacity management decisions, such as ward dimensioning, staffing decisions, patient-to-bed assignment and the development surgery schedules. Currently, predictions are typically solely based on the doctor’s estimate of the Expected Discharge Date (EDD). In this presentation, we propose two probabilistic models to combine the EDD with the LOS distributions retrieved from ERP data. Using the Poisson Binomial distribution and probabilistic convolution, we obtain the full census distribution. We apply our method on real world hospital data, and it turns out to be very accurate.

Keywords

Status: accepted


Back to the list of papers