Informs Annual Meeting 2017

MD06

INFORMS Houston – 2017

MD04

2 - Challenges in Modelling Time-dependent Transitions in Markov Cost-effectiveness Analysis Jiaru Bai, Binghamton University, School of Management, Binghamton, NY, 13902, United States, jiarub@uci.edu, L. Robin Keller We propose challenges in estimating Markov model parameters with limited clinical data. With limited data from registration trials, we developed a way to estimate time-dependent transition probabilities in a Markov model for cancer progression. The method is applied to cervical cancer data in a clinical study. 3 - Yardstick Competition for Service Systems Ozlem Yildiz, Assistant Professor, University of Virginia, Darden School of Business, Charlottesville, VA, United States, YildizO@Darden.virginia.edu, Nicos Savva, Tolga Tezcan We study whether an alternate pay-for-performance method can alleviate Emergency Department (ED) overcrowding through incentivizing socially-desired ED capacity levels, although the healthcare regulator does not know the capacity cost structure. Using yardstick competition, we propose a regulatory scheme that achieves this using the waiting time and arrival rate information of each ED. 4 - Optimal Treatment Model for Choosing Art Initiation Time Huaiyang Zhong, Stanford University, 119 Quillen Ct, Apt 217A, Stanford, CA, 94305, United States, hzhong34@stanford.edu, Xiaocheng Li, Margaret L. Brandeau HIV is now a chronic disease with the more and more availability of antiretroviral therapy (ART). WHO is advising that an infected individual should start ART as soon as the CD4 level drops to 500. However, because of diverse characteristics and different risk attitude of each patient, a personalized starting time of ART for each patient becomes very important. In this presentation, we present a new algorithm of Markov Decision Process (MDP) which will take into account of risk attitude and it will help solve the problem presented above. 320C Hospital Operations Management II Sponsored: Health Applications Sponsored Session Chair: Hui Zhang, University of Chicago, Chicago, IL, 60637, United States, corinnaz@uchicago.edu Chair: Thomas Best, tbest3@bsd.uchicago.edu 1 - A Queueing Framework for Evaluating Adoption of New Diagnostic Tests into Clinical Workflow with Application to Pulmonary Embolism in EDS Zhongjie Ma, Purdue University, West Lafayette, IN, 47906, United States, zhongjiema@purdue.edu, Pengyi Shi, H. Sebastian Heese, Jonathan Helm, Alice Marina Mitchell In medical research, new diagnostic tests are developed and evaluated solely on their efficacy in detecting a disease. However, ignoring the workload impact of introducing new tests into existing workflow can create barriers to adoption, particularly in busy units e.g. Emergency Department (ED). In collaboration with an ED physician, we develop a queueing framework for evaluating the workload impact of adopting new tests to bridge the gap between medical research and clinical workflow. We apply the framework to a new test, D-dimer, for detecting Pulmonary Embolism in EDs to understand what characteristics make adopting this test feasible and how to best integrate it into the ED workflow. 2 - Simulation-based Optimization to Improve Hospital Inpatient Placement under Capacity Strain Hui Zhang, University of Chicago, Chicago, IL, 60637, United States, corinnaz@uchicago.edu, Thomas Best, Anton Chivu, David O.Meltzer Many hospitals are confronted with capacity strain at their inpatient units resulting in delayed patient admissions, rejected admissions, and physician burnout. We develop a simulation-based optimization framework to improve hospital resource utilization using inpatient placement and overflow strategies. We use Discrete-Event Simulation approach and formulate an optimization model that maximizes hospital and physician revenue while accounting for various impact of patient overflow. We apply the model to a case study of hospitalist inpatient sector in University of Chicago Medical Center. Our findings inform hospital decision makers of how to address the capacity strain challenge. MD06

320A Sustainability and Social Responsibility in Operations Sponsored: Manufacturing & Service Oper Mgmt, Sustainable Operations Sponsored Session Chair: Sang Kim, Yale University, New Haven, CT, 06520, United States, sang.kim@yale.edu 1 - Ensuring Corporate Social and Environmental Responsibility through Vertical Integration and Horizontal Sourcing Adem Orsdemir, University of California Riverside, Anderson Hall, Riverside, CA, 92507, United States, orsdemiradem@gmail.com, Bin Hu, Vinayak V. Deshpande We investigate vertical integration as an alternative strategy for ensuring corporate social and environmental responsibility (CSER) at a supplier in a competitive setting. We show that when horizontal sourcing from a competitor is not feasible, NGOs’ tighter scrutiny may discourage a firm from vertically integrating with its supplier and becoming responsible. However, we also find that possibility of supplying a competitor and existence of demand externalities due to violation exposures may fundamentally change firms’ behaviors and CSER outcome. 2 - Can Brands Claim Ignorance? Unauthorized Subcontracting in Apparel Supply Chains Anna Saez de Tejada Cuenca, UCLA, Anderson School of Management, Los Angeles, CA, United States, anna.sdtc.1@anderson.ucla.edu, Felipe Caro, Leonard Lane The collapse of the Rana Plaza building in Bangladesh brought into focus the poor safety conditions faced by many workers in the apparel industry. A common way in which safety and environmental standards are violated is through unauthorized subcontracting. We analyze empirically some factors that can lead suppliers to outsource their production to third parties without their retailers’ knowledge. We use data provided by a supply chain manager that consists of over 30,000 orders, including 36% of subcontracted ones. Our results provide managerial insights to retailers on what factory and order characteristics increase the probability of unauthorized subcontracting, and how it can be prevented. 3 - Design of Public Warning Systems Saed Alizamir, Yale University, 339 Willow St., New Haven, CT, 06511, United States, saed.alizamir@yale.edu, Francis E. De Vericourt, Shouqiang Wang We study the design of public warning systems in a multi-period model. In each period, the sender (she) receives an imperfect signal about the true state of the world (dangerous or safe), and has to decide whether to warn the receiver to act. Depending on the true realization of the random event, the receiver updates his belief about sender’s credibility. The sender, therefore, has to dynamically manage her reputation over time, while also incentivizing the receiver to act on her warnings. We characterize the optimal warning policy, and gain insights into why it may sometimes be optimal for the sender to distort her signal. 320B Applications of Stochastic Models in Medical Decision Making Problems Sponsored: Health Applications Sponsored Session Chair: Saeideh Mirghorbani, University of Alabama, Tuscaloosa, AL, 35401, United States, smirghorbani@crimson.ua.edu 1 - The Effect of Adherence on the Anti-Hypertensive Therapy Plans in Patients with Type 2 Diabetes Saeideh Mirghorbani, University of Alabama, 1214, 12th St, Apt 17, Tuscaloosa, AL, 35401, United States, smirghorbani@crimson.ua.edu, Sharif Melouk, John Mittenthal The importance of adherence to the anti-hypertensive medications in patients with diabetes has been emphasized. Adherence is the extent to which the patient follows medical instructions and has been identified as the primary reason for suboptimal clinical benefits. In this research, we address the optimal antihypertensive therapy plans while considering patient adherence, the medication side-effects, and the cardiovascular complications. We develop a finite horizon, discounted Markov decision process in which patients transition through health states. The objective is to maximize the expected quality-adjusted life years. MD05

237