A Statistical Method for Predicting Inpatient Bed Demand  
Author Samuel R. Davis


Co-Author(s) Nasser Fard


Abstract The consequences of a mismatch in the supply and demand of hospital resources include non-clinical transfers and diversions, over-stretched resources, and under-utilized capacity. Predicting demand in advance can lead to achieving appropriate safety standards and reducing costs by adjusting staffing levels and patient flow protocols. This paper develops a flexible probability model of patient flow to estimate the probability mass function of future bed demand and is validated using operational data from a mid-sized community hospital. This approach improves upon previous work by providing a probability mass function of demand instead of a point estimate, using the exact surgery schedule instead of assuming a cyclic surgery schedule, and using patient-specific departure probabilities. The primary result of this work is a bed demand forecasting model with 5.5% average error that provides managers better information to optimize short-term adaptations to stochastic bed demand.


Keywords Bed demand prediction; service system reliability; adaptive staffing
    Article #:  24166
Proceedings ISSAT International Conference on Reliability and Quality in Design 2018
August 2-4, 2018 - Toronto, Ontario, Canada