Informs Annual Meeting 2017

MD03

INFORMS Houston – 2017

Monday, 4:30 - 6:00PM

2 - Real-time Bed Assignment Improves Patient Flow at Massachusetts General Hospital Andrew Martin Vanden Berg, Massachusetts Institute of Technology, Cambridge, MA, United States, avdberg@mit.edu, Aleida Braaksma, Bethany Daily, Peter Dunn, Retsef Levi, Ben Orcutt, Ana Cecilia Zenteno Assigning newly admitted patients to beds involves multiple constraints. Suboptimal patient-to-bed assignment may cause excessive patient wait times, crowding in upstream areas, and prolonged length-of-stays. We develop algorithms for real-time assignment of patients to beds, aiming at shortening patient wait times and decreasing upstream congestion. Through data-driven simulation, we quantify the benefits of the developed algorithms. Implementation of our results at Massachusetts General Hospital has resulted in significant decreases in patient wait times for beds, in addition to multiple other benefits. 3 - Risk-sensitive Stochastic Surgery Scheduling using CVAR Amirhossein Najjarbashi, University of Houston, 6377 Chevy Chase Dr, Apt 210, Houston, TX, 77057, United States, amirhossein.najjarbashi@gmail.com, Gino J.Lim This paper considers a daily surgery scheduling problem where surgical time requirements are stochastic with known probability distributions. A scenario- based mixed-integer linear programming formulation is proposed to minimize the conditional value-at-risk (CVaR) of overtime costs. Besides having desirable mathematical properties, applying CVaR enables hospital managers to adjust their approach toward the uncertainty using different probability levels. A set of benchmark instances is solved to evaluate the effectiveness of the proposed approach in providing low-cost and risk-averse schedules. 4 - Is More Information Always Better in Predicting Daily Surgical Case Volume? Joonyup Eun, Vanderbilt University Medical Center, 1211 21th Avenue South, Medical Arts Building, Nashville, TN, 37212, United States, joonyup.eun@vanderbilt.edu, Vikram Tiwari, Warren S. Sandberg We evaluate the tradeoff among two different dimensions of information - quality and quantity of information, by testing the accuracy of a series of data models to predict the daily surgical case volume. We retrieve detailed data of surgeons’ working patterns from several information sources. We find that greater amount of information may not improve prediction accuracy if the additional information is prone to errors; however, deploying analytical data treatment strategies to ameliorate those errors improves information quality and results in better surgical case volume predictions. 310C 4:30- 5:15 Gurobi / 5:15 - 6:00 Princeton Consultants Inc. Invited: Vendor Tutorial Invited Session 1 - Optimization with R Renan Garcia, Gurobi Optimization, Houston, TX, United States, garcia@gurobi.com Are you interested in leveraging mathematical programming technology from within R? This tutorial will survey some of the more popular optimization-related packages in the vast R ecosystem. This includes an overview of the Gurobi R package, along with detailed examples of how to build the optimization models you are looking to solve. 2 - Analytics Model Review and Validation Irvin J.Lustig, Princeton Consultants, 25 Sylvan Way, 2 Research Way, Short Hills, NJ, 07078, United States, irv@princeton.com Acting as an independent third party, Princeton Consultants reviews analytics models and how they are deployed in a business. Through our Advanced Analytics Model Review and Validation service, we ask questions such as: What is a correct model? What data is being integrated and how? How are solutions published and used in the business? How sensitive are the answers to the inputs? Did the implemented model reflect the intentions of the practitioner? In this tutorial, Irv Lustig will illustrate the importance of addressing these questions in the context of deploying advanced analytics models in practice. MD03

MD01

310A Decision Analysis Awards Sponsored: Decision Analysis Sponsored Session Chair: Jason Merrick, Virginia Commonwealth University, jrmerric@vcu.edu 1 - DAS Student Paper Award Emanuele Borgonovo, Bocconi University, Bocconi University, Via Sarfatti 25, Milano, 20136, Italy, emanuele Borgonovo@unibocconi.it, John Hahn The Student Paper Award is given annually to the best decision analysis paper by a student author, as judged by a panel of the Decision Analysis Society of INFORMS. Students who did not complete their PhD prior to May 1, 2016 are eligible for this year’s competition. 2 - DAS Publication Award Vicki Bier, University of Wisconsin, Dept of Ind. Eng. Mechanical Eng Bldg., 1513 University Avenue, Madison, WI, 53706, United States, bier@engr.wisc.edu This award is given annually to the best decision analysis article or book published in the second preceding calendar year (i.e. calendar year 2015for consideration in 2017). The intent of the award is to recognize the best publication in “decision analysis, broadly defined.” This includes, but is not limited to, theoretical work on decision analysis methodology (including behavioral decision making and non-expected utility theory), descriptions of applications, and experimental studies. 3 - DAS Practice Award Gregory L. Hamm, Stratelytics, LLC, 933 Taylor Avenue, Alameda, CA, 94501, United States, ghamm@strts.com The Decision Analysis Practice Award is awarded to the best example of decision analysis practice as judged by the Decision Analysis Practice Award Committee. The purpose of the award is to publicize and encourage outstanding applications of decision analysis practice. We will present the finalists and this year’s winner. 4 - DAS Ramsey Medal J. Eric Bickel, The University of Texas at Austin, Graduate Program in ORIE, 1 University Station, C2200, Austin, TX, 78712-0292, United States, ebickel@mail.utexas.edu The Ramsey Medal of the Decision Analysis Society is awarded for distinguished contributions in decision analysis. Distinguished contributions can be internal, such as theoretical and procedural advances in decision analysis, or external, such as developing or spreading decision analysis in new fields. We will introduce the 2017 Ramsey Medal winner, followed by a presentation by the winner.

MD02

310B Managing Surgical and Inpatient Capacity Sponsored: Health Applications Sponsored Session

Chair: Joonyup Eun, PhD, Vanderbilt University Medical Center, 1211 21st Avenue South, Medical Arts Building, Nashville, TN, 37212, United States, joonyup.eun@vanderbilt.edu Co-Chair: Vikram Tiwari, Vanderbilt University Medical Center, Nashville, TN, 37221, United States, vikram.tiwari@vanderbilt.edu 1 - A Randomized Controlled Trial of a Predictive Modelling-based Scheduling System Panagiotis Kougias, MD, Baylor College of Medicine, Houston, TX, United States, pkougias@bcm.edu We randomly scheduled cases in a single operating room by predicting their length using either their historical mean (HM) duration or a regression modeling (RM) system that took into account operative and patient characteristics. Mean imprecision in predicting the end of operative day was higher with the HM approach (30.8 vs. 7.2 minutes, p=0.024). RM was associated with higher throughput (379 vs. 356 cases scheduled over the course of the study, p=0.04) and lower overutilization rate (34 vs. 51% of days with overutilization, p = 0.004). Mean length of overutilization (58 vs. 93 minutes, p=0.002) and undertulization (42 vs. 65 minutes, p=0.013) were superior in the RM arm.

235