2015 Informs Annual Meeting

SA36

INFORMS Philadelphia – 2015

3 - Multi-Server Queues with Impatient Customers as Level- Dependent QBDs with Applications in Healthcare Amir Rastpour, University of Alberta, Alberta School of Business, PhD office, Edmonton, AB, T6G 2R6, Canada, amir.rastpour@ualberta.ca, Burhaneddin Sandikci, Armann Ingolfsson We investigate the use of level-dependent quasi-birth-death (QBD) processes to analyze priority queues with impatient customers, such as emergency departments where patients are triaged into priority classes and some patients leave without being seen. We report numerical results and discuss algorithm performance (accuracy and speed). 4 - Optimal Cut-off Points for RNA-based Testing to Minimize the Transfusion-transmitted Infection Risk Hrayer Aprahamian, PhD Student, Virginia Tech, Dept of ISE, Blacksburg, VA, 24061-0118, United States of America, ahrayer@vt.edu, Ebru Bish, Douglas Bish The safety of blood products, in terms of being free of infectious agents (e.g., human immunodeficiency virus, hepatitis viruses), is essential. We develop a novel mathematical model to determine the optimal cut-off points for RNA-based individual and pooled screening tests, considering all recognized and emerging infections that can be transmitted through the use of blood products. Using real data, we show that our model improves upon current practices.

4 - Evaluating Different Policies: A Real Life Operating Room Scheduling Problem Elvin Coban, Ozyegin University, Cekmekoy, Istanbul, Turkey, elvin.coban@ozyegin.edu.tr, Gulsah Alper We study a real life operating room scheduling problem using a dataset from a leading hospital in Turkey. We solve the daily and weekly scheduling problems by a mixed integer linear programming model. Various objective functions and performance metrics are analyzed including minimizing the waiting time of patients while maximizing fairness. We examine surgery delays and incorporate possible delays in surgery durations. We also propose a method to compute robust operating room schedules. SA36 36-Room 413, Marriott Humanitarian Applications I Sponsor: Public Sector OR Sponsored Session Chair: Burcu Balcik, Ozyegin University, Nisantepe Mah. Orman Sok. Cekmekoy, Istanbul, Turkey, burcu.balcik@ozyegin.edu.tr 1 - The Needs Assessment Routing Problem in Humanitarian Relief Burcu Balcik, Ozyegin University, Nisantepe Mah. Orman Sok. Cekmekoy, Istanbul, Turkey, burcu.balcik@ozyegin.edu.tr In the immediate aftermath of a disaster, it is important for relief agencies to develop accurate estimates about the effects of the disaster in the affected region. Since assessments must be completed quickly, it may not be possible to visit each site in the affected region to collect information. We present mathematical models and solution approaches to support site selection and routing decisions of the rapid needs assessment teams. We present a case study to illustrate our approach. 2 - Effective Response to Disable and Elderly Populations in Short-notice Disasters Jacqueline Griffin, Assistant Professor, Northeastern University, 334 Snell Engineering Center, 360 Huntington Ave, Boston, MA, 02125, United States of America, ja.griffin@neu.edu, Rana Azghandi We develop a mixed integer programing model to simultaneously account for the different protection strategies for the elderly and disabled population in short- notice disasters. The modeling poses a split-delivery vehicle routing problem with time windows and multiple uses of heterogeneous vehicles. We examine the effect of multiple objectives for this disaster response application. Moreover, the value of cooperation among neighboring jurisdictions as compared with greedy policies is examined. 3 - A Dynamic Model for Disaster Response Considering Prioritized Demand Points Gina Galindo, Dr., Universidad del Norte, Km 5 Antigua Via a Puerto Colombia, Barranquilla, Colombia, ggalindo@uninorte.edu.co, Daniel Rivera This research addresses the problem of distributing relief supplies after the occurrence of a natural disaster. We develop a dynamic model to define an action plan to serve demand, while prioritizing the response according to the level of urgency of demand points. Our model considers capacity constraints and dynamic priorities. To evaluate its applicability, we use a case study of a flood occurred in Colombia. We also test the solvability of our model for large instances of our problem. 4 - The Role of Media Exposure on Humanitarian Donation and Coordination Mahyar Eftekhar, Arizona State University, P.O. Box 874706, Tempe, AZ, 85287, United States of America, eftekhar@asu.edu, Luk Van Wassenhove, Hongmin Li, Scott Webster Despite their resource and financial limitations and despite the considerable level of demand uncertainty they face, Humanitarian Organizations (HOs) do not typically share resources. Considering the impact of media exposure, our study unveils the conditions in which humanitarians will coordinate. This paper contains both empirical and analytical modeling.

SA35 35-Room 412, Marriott Joint Session HAS/MSOM-Healthcare: Health Care Operations Sponsor: Health Applications & MSOM Sponsored Session

Chair: Nilay Argon, University of North Carolina, Department of Statistics and Operations, Chapel Hill, NC, 27599, United States of America, nilay@unc.edu Co-Chair: Serhan Ziya, Associate Professor, UNC Department of Statistics & Operations Research, 356 Hanes Hall, CB#3260, Chapel Hill, NC, 27599 - 32, United States of America, ziya@unc.edu 1 - Scale and Skill Mix Efficiencies in Nursing Home Staffing Ger Koole, VU University Amsterdam, De Boelelaan 1081a, Amsterdam, Netherlands, ger.koole@vu.nl, Dennis Moeke, Lineke Verkooijen Care workers account for a significant proportion of the total health expenditure in nursing homes and are by far the largest controllable resource. Therefore determining the appropriate number and type of care workers required plays an important role in the search for more efficiency. This study provides insights in how and why scale of scheduling and blending tasks of different qualification levels effect the number and type of staff required to meet the preferences of nursing home residents. 2 - Myopic Scheduling of Jobs with Decaying Value with Applications in Patient Scheduling Neal Master, Stanford University, 350 Serra Mall, Stanford, CA, 94305, United States of America, nmaster@stanford.edu, Carri Chan, Nicholas Bambos In healthcare settings, delays in receiving treatment can result in worse outcomes for patients. We introduce a clearing model in which the reward generated by completing service for an individual job decays over time. Because computing an optimal policy for such a model is computationally intractable, we focus on a number of myopic heuristics. We provide performance guarantees for each heuristic and use simulation to gain further insight into patient scheduling problems. 3 - Wait Time Announcements at Hospital Emergency Departments Zhankun Sun, Haskayne school of business, University of Calgary, We study a multiclass multiserver priority queue with delayed feedback to predict the wait time for low-priority patients to be seen by a physician for the first time after triage in ED. We model the patients reassess process and develop a procedure to predict the state-dependent wait time based on an busy-period analysis. With a case study at the four major hospitals in the Calgary area we illustrate the performance of our wait time predictions. 2500 University Drive NW, Calgary, AB, Canada, zhankun.sun@haskayne.ucalgary.ca, Marco Bijvank

49