Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

MD60

2 - Optimizing Population Screening for Infectious Diseases. The Case of Sleeping Sickness Control in Congo Harwin de Vries, INSEAD, Boulevard de Constance, Fontainebleau, 77210, France Population screening by mobile units is crucial to control several infectious diseases. We consider the following planning problem: given a set of populations at risk, the expected evolution of the epidemic in these populations, and a fixed number of mobile units, which villages should be screened when? We present descriptive models for the development of the burden of disease over time which take screening explicitly into account, use these to develop and analyze several classes of planning policies, and numerically analyze them in the context of the control of the HAT disease in D.R. Congo. 3 - Individual Wait Time Estimation in the Organ Allocation System Ana Maria Estrada Gomez, Georgia Institute of Technology, Atlanta, GA, United States, Kamran Paynabar When a patient is offered an organ, he/she needs to decide whether to accept or decline it. This complex decision depends in part on the estimated wait time until the next organ offer and that organ’s quality. Using a match-run database, we develop methods to predict the wait time for an organ (liver, kidney, lung, heart) offer of a desired quality, given the patient’s characteristics and compare our estimates to those provided by UNOS. 2 - Optimizing Population Screening for Infectious Diseases. The Case of Sleeping Sickness Control in Congo oris Van de Klundert, Albert P M Wagelmans Population screening by mobile units is crucial to control several infectious diseases. We consider the following planning problem: given a set of populations at risk, the expected evolution of the epidemic in these populations, and a fixed number of mobile units, which villages should be screened when? We present descriptive models for the development of the burden of disease over time which take screening explicitly into account, use these to develop and analyze several classes of planning policies, and numerically analyze them in the context of the control of the HAT disease in D.R. Congo. 3 - Individual Wait Time Estimation in the Organ Allocation System Joel Sokol, Pinar Keskinocak When a patient is offered an organ, he/she needs to decide whether to accept or decline it. This complex decision depends in part on the estimated wait time until the next organ offer and that organ’s quality. Using a match-run database, we develop methods to predict the wait time for an organ (liver, kidney, lung, heart) offer of a desired quality, given the patient’s characteristics and compare our estimates to those provided by UNOS. 4 - A Scan Statistic for Disease Detection Dave Goldsman, Georgia Institute of Technology, School of ISyE, Atlanta, GA, 30332-0205, United States, Ignacio Erazo Consider a square containing a number of uniformly distributed points. We use analytical methods and Monte Carlo simulation to obtain the distributions of certain scan statistics, for instance, the maximum number of points contained in a smaller square as it sweeps through the original square. These distributional results lead to statistical tests to detect the occurrence of a significant event such as a disease outbreak. Several examples illustrate our method’s efficacy, statistical power, and ease of use. n MD59 West Bldg 102A Methods for Decision Making in Healthcare Sponsored: Health Applications Sponsored Session Chair: Vishal Ahuja, Southern Methodist University, Dallas, TX, 75275, United States 1 - Using a Deep Learning Algorithm to Quantify Common Barriers to Discharge Taghi Khaniyev, MIT Sloan School of Management, Cambridge, Hospital congestion is caused among other factors by an imbalance of the timing of patient admissions and discharges. We implemented a real-time clinical support tool to predict next-day’s surgical patient discharges based on previous patients’ recovery care paths at a large academic medical center. Further, we analyzed mismatches between predicted and actual discharges to identify areas for improvement in the discharge process and to enhance the accuracy of the tool. Results show that the top 10 patients predicted to leave are discharged 81% of the time within 24 hours. Mismatches occurred without clinical reason 45% of the time and were due to clinical practice variation 25% of the time.” MA, United States, Dominic Breuer, Kyan Safavi, Ana Cecilia Zenteno, Mark Seelen, Jonathan Zanger, Bethany Daily, Peter Dunn, Retsef Levi

2 - ”Wait and then Hurry Up” – the Effect of Flexible Service Rates in Appointment Scheduling Systems Aditya Jain, Baruch College, Zicklin School of Business, 55 Lexington Ave, Suite 9-240, New York, NY, 10010, United States, William P. Millhiser Recent empirical evidence suggests that ambulatory care healthcare workers vary service rates in response to time of day and queue length. In this paper, we explore a theoretical justification of this behavior using trade-off between waiting and processing costs and analyze its implications for established appointment systems, and how they should be adapted. 3 - A New Patient Trajectory Simulation Method for Staffing in Care Coordination Programs Ekin Koker, University of Massachusetts, Mechanical and Industrial Engineering, 160 Governors Drive, Amherst, MA, 01003, United States, Rebecca Castonguay, Hari Balasubramanian, Zachary Martinez, Aaron Truchil Care Coordination programs intended to improve outcomes and lower cost of care for patients with complex health and social needs have proliferated in recent years. Staff encounters with patients in these programs can be modeled as a trajectory, each encounter marked with a staff type, duration and a timestamp from enrollment date. Using encounter data from a care coordination program, we simulate different realizations by sampling patient trajectories randomly without replacement, allowing for a constant and a Poisson patient arrival rate. We present our findings and offer insights for healthcare organizations looking to start care coordination programs. 4 - Optimization Approaches for Concussion Management Decisions Gian-Gabriel Garcia, Mariel Sofia Lavieri We apply optimization for concussion management decisions. n MD60 West Bldg 102B Scheduling and Resource Allocation in Health Systems Sponsored: Health Applications Sponsored Session Chair: Pinar Keskinocak, Georgia Institute of Technology, Atlanta, GA, 30332, United States Co-Chair: Seyma Guven Kocak, Georgia Tech, Atlanta, GA, United States 1 - Staff Planning with Cost Estimation and Optimization Kumar Rajaram, Universityof California-Los Angeles, Anderson School, 110 Westwood Plaza, Los Angeles, CA, 90095-1481, United States, Sandeep Rath We consider the staff planning problem of anesthesiologists which consists of determining how many anesthesiologists have to be on regular duty and on call each day. We model this problem as a two-stage integer stochastic dynamic program. We use structural estimation to assess the implicit costs in this model. We develop a solution algorithm, and test the model with real data from the UCLA medical center. 2 - Home Health Care Routing and Scheduling with Continuity of Care This work addresses a real-world home health care routing and scheduling problem (HHCRSP) faced by a home care agency in the U.S. In home health care scheduling, there is a desire to retain consistency with respect to the nurses servicing each patient, which is referred to as continuity of care. In order to handle this requirement, we propose a dynamic approach and present two different methods to solve the problem on a daily basis: a MIP with approximations and a variant of a petal heuristic. We aim to quantify and control the deviation from the existing schedule in place, so that some of the assignments are retained in the new schedule. We discuss the performance and computational efficiency of these methods. Seyma Guven-Kocak, Georgia Institute of Technology, 1235 Parkway Circle N, Atlanta, GA, 30340, United States, Aliza R. Heching, Pinar Keskinocak, Alejandro Toriello

235