Healthcare Associated Infections (HAIs) are infections acquired by patients admitted in healthcare facilities. These infections can be contagious or even worse, fatal, which has prompted the medical community to put guidelines in place to isolate patients who may pose a risk of spreading or be susceptible to an infection(s). These isolation practices also have the unwanted consequence of constraining the options for inpatient bed assignments since infected patients cannot be assigned to a room where uninfected patients reside. Historically, hospitals in the United States have been built with single and double bedrooms to provide inpatient care. As long as demand for a bedroom is less than the number of bedrooms in the hospital, bed assignment is a trivial process. Difficulties for determining bed assignments occur when hospital units operate at full or nearly full bed utilization and must continue to admit new patients. When this occurs, the units' administrators must determine when to admit new patients, whether current patients need to be discharged to make room for new more critical patients, or if there is a need to exchange the rooms of already admitted patients (i.e., internal movement). These decisions are complicated by the limited bedroom capacity (number of rooms and occupancy threshold in rooms) and by the need to implement isolation guidelines necessary to prevent and contain the occurrence of HAIs. This study presents two optimization models to suggest how to accommodate admitted and incoming patients in a hospital unit to satisfy all isolation requirements, while simultaneously maximizing the total criticality of patients admitted into the unit and minimizing the number of internal movements. These models provide bed assignment recommendations based on available bed demand and patient characteristic information for a current planning period first for current (known) demand, and then considering uncertain future demand. Additionally, we explore the use of the first model to suggest a methodology for determining the number of single and double bedrooms in a hospital unit as well as the stockpile of spare resources necessary to ensure a desired service level for inpatient hospitalization demand, when isolation requirements are considered. Finally, the second model which considers future demand is explored to examine the effect on bedroom assignments incurred from the number of periods used in the planning horizon given a small subset of hypothetical, yet realistic hospital data.

Library of Congress Subject Headings

Hospitals--Administration--Computer simulation; Hospital patients--Classification; Hospital size--Computer simulation; Nosocomial infections--Prevention

Publication Date


Document Type


Department, Program, or Center

Industrial and Systems Engineering (KGCOE)


Proaño, Rubén


Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works. Physical copy available through RIT's The Wallace Library at: RA971 .C44 2013


RIT – Main Campus

Plan Codes