Informs Annual Meeting 2017

SB43

INFORMS Houston – 2017

SB42

4 - A Two-Stage Vehicle Routing Algorithm Applied to Disaster Relief Logistics after the 2015 Nepal Earthquake Stephanie Allen, The State University of New York at Geneseo, New York, NY, United States, stephanie.a.allen95@gmail.com, Caroline Haddad After the 2015 Nepal Earthquake, the Himalayan Disaster Relief Volunteer Group distributed supplies to affected areas. We model the organization’s operations as a vehicle routing problem using Fisher and Jaikumar’s two-stage method, which allocates locations to vehicles via an assignment integer program and then uses heuristics to route the vehicles. Our results illustrate the open nature of the VRP and the computational necessity of heuristics. 5 - Contract Tracing Protocol Chungjae Lee, Georgia Institute of Technology, Atlanta, GA, United States, clee384@gatech.edu, Jason Bermudez, Kevin Desprez, Alex Kehres, Leah Patterson, Suphaphat Petlerkwang, Yuntong Zhu, Pinar Keskinocak, Brian Gurbaxani Contact tracing is a protocol that prevents disease transmission by identifying disease-carriers. The Center for Disease Control and Prevention (CDC) suspects that the protocol is improperly estimating the number of contacted individuals and incurring excessive cost. Mathematical models and a decision tool are developed to design cost-effective protocols with quantitative justification for respiratory diseases. Through the tool, the CDC can reduce cost at most by $360,000 per protocol. 352F Health Care, Modeling and Optimization Contributed Session Chair: Xiaoli Duan, Northeastern University, Boston, MA, United States, duan.xi@husky.neu.edu 1 - Improving Outpatient Scheduling with Simulation Michelle Rose Hribar, Oregon Health & Science University, Portland, OR, United States, hribarm@ohsu.edu, Clinicians today face increased patient loads, decreased reimbursements and negative productivity impacts from using electronic health records (EHR), but have no guidance on how to improve clinic efficiency. This study evaluates novel scheduling strategies using simulation and EHR data before their implementation in outpatient ophthalmology clinics. Key findings from this study are: 1) EHR timestamp data in simulation models represents clinic workflow, 2) simulations provide critical information for schedule creation and resource allocation, 3) scheduling long appointments last reduces patient waits, and 4) new schedule implementation challenges limit efficiency improvements. 2 - Establishing Staffing Levels and Surge Planning for the Emergency Department of the Children’s Hospital of Pittsburgh using Simulation Kiatikun Louis Luangkesorn, University of Pittsburgh, 1028 Benedum Hall, 3700 Ohara St, Pittsburgh, PA, 15261, United States, lluangkesorn@pitt.edu, Anna Svirsko, Bryan A. Norman, Jayant Rajgopal, John Rossi An urban hospital emergency department experiences high variability in arrivals over the course of a day, from day to day, and from season to season. To handle this variable demand, Children’s Hospital uses both a standard staffing level and a surge team for when the standard staffing level is not able to keep up with the patient demand. We present a simulation model to jointly analyze staffing levels and the surge plan in this environment. This simulation is used to evaluate the effects of staffing on ED length of stay and the need to implement the surge plan over the course of the year. 3 - Ophthalmology Benchmarking using Data Envelopment Analysis Timothy R.Anderson, Portland State University, Mail Code ETM, P.O. Box 751, Portland, OR, 97201, United States, tim.anderson@pdx.edu, Daria Spatar A benchmarking study of the electronic medical records adoption and usage by ophthalmology practices was conducted using data envelopment analysis. 4 - Agent-based Models of Opioid Use Spread Xiaoli Duan, Northeastern University, Boston, MA, United States, duan.xi@husky.neu.edu, James Benneyan, Iulian Ilies We developed several agent-based models of the spread of opioids abuse to help understand and mitigate this worsening national epidemic. Models varied in stochasticity, scale (individual to region), and topology (geographical to social) and were tested on several sociodemographic populations. Results indicated that different models work better in different settings and can help identify the magnitudes and combinations of interventions and investments (e.g., reduced prescription availability, education initiatives, or increased recovery services) necessary to effectively reduce the epidemic. SB41 Sarah Read-Brown, Leah Reznick, Lorinna Lombardi, Mansi Parikh, Winston Chamberlain, Thomas Yackel, Michael F. Chiang

360A Journal Editors Session Invited: Invited OR and Advanced Manufacturing Invited Session 1 - Journal Editors Panel

Satish Bukkapatnam, Texas A&M.University, 3131 TAMU, 4020 Emerging Technologies Building, College Station, TX, 77843, United States, satish@tamu.edu This session will feature three panelists representing prominent journals in the manufacturing area to discuss opportunities to communicate research at the intersection of advanced manufacturing and operations research in these journals. 2 - Focus Issue Editor, IIE Transactions on Design and Manufacturing Shiyu Zhou, University of Wisconsin-Madison, 3254 Mech Eng Bldg, 1513 University Avenue, Madison, WI, 53706-1572, United States, szhou@engr.wisc.edu 3 - Department Editor, Production and Operations Management Chelliah Sriskandarajah, Texas A&M.University, 320q Wehner Building, 4217 Tamu, College Station, TX, 77843-4217, United States, chelliah@mays.tamu.edu 4 - Associate Editor, Journal of Manufacturing Science and Engineering Satish Bukkapatnam, Texas A&M.University, 3131 TAMU, 4020 Emerging Technologies Building, College Station, TX, 77843, United States, satish@tamu.edu 360B Advances in Healthcare Analytics Sponsored: Public Sector OR Sponsored Session Chair: Van-Anh Truong, Columbia University, New York, NY, 10027, United States, vt2196@columbia.edu 1 - A Fluid Model for an Overloaded Bipartite Queueing System with Heterogeneous Matching Utilities Yichuan Ding, UBC, 2053 Main Mall, Vancouver, BC, V6T1Z2, Canada, daniel.ding@sauder.ubc.ca, S. Thomas McCormick, Mahesh Nagarajan We consider a bipartite queueing system (BQS) with multiple types of servers and customers, where different customer-server combinations may generate different utilities. Whenever a server is available, it serves the customer with the highest index, which is the sum of a customer’s waiting index and the matching index. We develop a fluid model to approximate the behavior of such a BQS system, and show that the fluid limit process can be computed over any finite horizon. We prove that the fluid limit process converges to a unique steady state, which can be efficiently computed. We illustrate the use of our machinery by analyzing the public housing assignment in the city of Pittsburgh using a real data set. 2 - Learning Personalized Treatment Regimes from Observational Data Nathan Kallus, Cornell University, New York, NY, United States, kallus@cornell.edu In this talk I will describe several recent advances in learning personalized decision policies from observational data and its application to personalized medicine using electronic health records, including new methods, problems, and applications. 3 - Conspicuous by its Absence: Diagnostic Expert Testing under Uncertainty Tinglong Dai, Assistant Professor, Johns Hopkins University, Baltimore, MD, 21202, United States, dai@jhu.edu, Shubhranshu Singh Motivated by prevalent—albeit little explored—under-provision in the US healthcare setting, we study a scenario in which costly diagnostic testing is needed to acquire customers’ true conditions and the expert’s accuracy of diagnosis is unknown to customers. Our work is the first to formally link diagnostic uncertainty and information asymmetry. SB43

49