Operations Research for Occupancy Modeling at Hospital Wards and Its Integration into Practice

Abstract

Medical and technological advancements are contributing to increase healthcare expenditures and increase numbers of hospitalized patients (Chernew and Newhouse 2012), while at the same time the length of stay (LoS) for these patients decreases. However, healthcare expenditures are still rising (OECD). Society calls for improved cost effective healthcare delivery, which puts pressure on available N.M. van de Vrugt ( )

Integral Capacity Management & Planning in Hospitals

“…An important decision at this level is how many staffed beds to assign to individual wards. From a logistical viewpoint, larger wards will result in economies of scale, leading to a higher occupancy with an acceptable blocking probability (see [226]). For this decision, the trade-off between medical and logistical perspectives should be considered.…”

“…On a strategic level, the hospital board may set performance targets for the hospital. In hospital wards, the logistical performance indicators used are often the bed census or occupancy, with occupancy measurable in many different ways [226]. Setting high occupancy targets for wards will, certainly for smaller wards, result in deferring more patients to other wards or hospitals.…”

“…The authors in [226] use a mixed integer programming approach to develop a tactical schedule for a weekday ward. Weekday wards admit elective patients only on weekdays and close on the weekends.…”

“…In accordance with the list of factors in [226], at the start of this project we commissioned a steering group consisting of all stakeholders in this problem: a neurologist, an internal medicine specialist, an administrator from the patient admission scheduling office, and all involved ward managers. The hospital management made this steering group responsible for finding a solution for the over-and under-capacity of the wards, and included one organizational advisor and a healthcare logistics advisor in the steering group.…”

“…The effects of intervention 1 were analyzed for several scenarios using an M (t)/M (t)/s/s queue [226]. The conclusion of this analysis was that the AMU would not be beneficial for the hospital's case-mix, and the steering group discarded this intervention as well.…”