Roumani, Yazan
(2013)
Modeling Patient Flow in a Network of Intensive Care Units (ICUs).
Doctoral Dissertation, University of Pittsburgh.
(Unpublished)
Abstract
Beginning in 2012, the Department of Health and Human Services (HHS) started adjusting payment for specific conditions by 30% for hospitals with 30-day patient readmission rates higher than the 75th percentile (HHS.gov, 2011). Furthermore, starting in 2013, HHS requires hospitals to publish their readmission rates (HHS.gov, 2011). It is also estimated that by 2013, healthcare expenditures in the United States will account for 18.7% of the Gross Domestic Product (GDP) (Centers of Medicare and Medicaid Services and US Bureau of Census, 2004). Yet the US healthcare system still suffers from congestion and rising costs as illustrated by hospital congestion.
One way to reduce congestion and improve patient flow in the hospital is by modeling patient flow. Using queueing theory, we determined the steady state solution of an open queueing network, while accounting for instantaneous and delayed feedback. We also built a discrete event simulation model of patient flow in a network of Intensive Care Units (ICUs), while considering instantaneous and delayed readmissions, and validated the model using real patient flow data that was collected over four years. In addition, we compared several statistical and data mining techniques in terms of classifying patient status at discharge from the ICU (highly imbalanced data) and identify methods that perform the best.
Our work has several contributions. Modeling patient flow while accounting for instantaneous and delayed feedback is considered a major contribution, as we are unaware of any patient flow study that has done so. Validating the discrete event simulation model allows for the implementation and application of the model in the real world by unit managers and administrators. The simulation model could be used to test different scenarios of patient flow, and to identify optimal resource allocation strategies in terms of number of beds and/or staff schedules in order to maximize patient throughput, reduce patient wait time and improve patients’ outcome. Moreover, identifying high risk patients who are more likely to die in the ICU ensures that those patients are receiving appropriate and timely care, so their risk of death is reduced.
Share
| Citation/Export: |
|
| Social Networking: |
|
Details
| Item Type: |
University of Pittsburgh ETD
|
| Status: |
Unpublished |
| Creators/Authors: |
|
| ETD Committee: |
|
| Date: |
2 July 2013 |
| Date Type: |
Publication |
| Defense Date: |
11 April 2013 |
| Approval Date: |
2 July 2013 |
| Submission Date: |
20 May 2013 |
| Access Restriction: |
1 year -- Restrict access to University of Pittsburgh for a period of 1 year. |
| Number of Pages: |
115 |
| Institution: |
University of Pittsburgh |
| Schools and Programs: |
Joseph M. Katz Graduate School of Business > Business Administration |
| Degree: |
PhD - Doctor of Philosophy |
| Thesis Type: |
Doctoral Dissertation |
| Refereed: |
Yes |
| Uncontrolled Keywords: |
ICU, feedback, readmission, network |
| Date Deposited: |
02 Jul 2013 13:49 |
| Last Modified: |
15 Nov 2016 14:12 |
| URI: |
http://d-scholarship.pitt.edu/id/eprint/18766 |
Metrics
Monthly Views for the past 3 years
Plum Analytics
Actions (login required)
 |
View Item |
![[feed]](/images/feed-icon-32x32.png){kind=icon}