Informs Annual Meeting 2017

MD16

INFORMS Houston – 2017

MD14

compete in a contest whose feasibility is uncertain. They learn about feasibility from their actions as well as from the progress of their competitors. A contest designer is interested in maximizing the efforts that agents exert, and utilizes the contest’s information disclosure mechanism as a lever to that end. We examine several information disclosure designs that a contest designer can utilize and we compare their properties in terms of the designer’s objective and the contestants’ welfare. 2 - Referral Priority Program: Leveraging Social Ties via Operational Incentives Luyi Yang, Johns Hopkins University, Baltimore, MD, 21202, United States, luyi.yang@jhu.edu, Laurens G. Debo The referral priority program, an emerging business practice adopted by a growing number of technology companies that manage a waitlist of customers, enables existing customers on the waitlist to gain priority access if they successfully refer new customers to the waitlist. We study the effectiveness of such a scheme as a marketing tool for customer acquisition and an operational approach for waitlist management. 3 - Competitive Equilibria in two Sided Matching Markets with Non- quasilinear Utility Functions Azarakhsh Malekian, Rotman School of Management, 105 Saint George, malekian, Toronto, ON, 02143, Canada, azarakhshm@gmail.com We consider two sided matching markets consisting of agents with non- transferable utilities; agents from the opposite sides form matching pairs (e.g., buyers-sellers) and negotiate the terms of their math which may include a monetary transfer. Competitive equilibria are the elements of the core of this game. We present the first combinatorial characterization of competitive equilibria that relates the utility of each agent at equilibrium to the equilibrium utilities of other agents in a strictly smaller market excluding that agent; thus automatically providing a constructive proof of existence of competitive equilibria in such markets. 4 - Choosing the Right Campaign Mode in Crowdfunding Simone Marinesi, Wharton, 201 S.25th Street, Apt 621, Philadelphia, PA, 75003, United States, marinesi@wharton.upenn.edu, Karan Girotra We compare the two most popular campaign modes employed by reward-based crowdfunding platforms: All-or-Nothing, in which the entrepreneur (creator) collects the money pledged by consumers (backers) only if the project is fully funded, and Flexible Funding, in which the entrepreneur collects the money even when the project is partially funded. A comparison between the two campaign modes reveals that an entrepreneur should choose a Flexible Funding campaign for projects with a low budget, high market prospects, or if capital markets are particularly accessible. 332F Inventory Management Contributed Session Chair: Yang Zhou, Huazhong University of Science and Technology, Wuhan, China, m201673799@hust.edu.cn 1 - An Analysis of Production Process Considering Yield Distribution Sungyong Choi, Assistant Professor, Yonsei University, College of Government and Business, 1 Yoneseidae-gil, Wonju, 26493, Korea, Republic of, sungyongchoi@gmail.com, Sumin Jeon, Jinmin Kim, Kwangtae Park We study a binomial yield production process using a newsvendor approach by incorporating underage and overage costs. We provide a reformulation of the original model to convert the discrete and exact model to the continuous and approximate model via normal approximation. We conduct comparative static analysis for model parameters in the approximate model and derive the monotone properties of the optimal solution. All of the analytical results are consistent with our insights and supported by economic explanations. Our numerical study with sample-based optimization indicates that the approximate model is close to the exact model in most of the real world examples including some limiting cases. 2 - Single Period Inventory Problem under Differentiated Service Levels using a Responsive Allocation Policy Pablo Escalona, Universidad Técnica Federico Santa María, Av Espana 1680, Valparaiso, 2390123, Chile, pablo.escalona@usm.cl We study the rationing of a single order quantity among multiple customer classes under stochastic demands and differentiated service levels using a responsive allocation policy when demand is made during the lead time. The problem is formulated as a nonlinear problem with chance constraints, for which we exploit the geometrical properties of the model, to propose a decomposition technique to solve it. Although the heuristic does not guarantee an optimal solution, our computational experiments have shown that it provides good-quality solutions that, under a mild assumption, are epsilon-optimal. MD16

332D Mathematical Models and Empirical Methods for Improving Healthcare Operations Sponsored: Manufacturing & Service Oper Mgmt, Healthcare Operations Sponsored Session Chair: Goh Joel, jgoh@hbs.edu 1 - Maximizing Intervention Effectiveness Brian Rongqing Han, Marshall School of Business, USC, Los Angeles, CA, United States, rongqing.han.2019@marshall.usc.edu, Vishal Gupta, Song-Hee Kim In the emerging paradigm of evidence-based decision making in healthcare and social science, policymakers often face challenges when deciding whom to receive an intervention that has previously been proven effective. With heterogeneous intervention effects and no access to raw experimental data, who should be selected for intervention? We propose a novel robust optimization framework that uses only summary statistics from a published paper to select a subset of the candidate population to maximize intervention effectiveness. We assess the practical significance of our method using data from a large teaching hospital who Alireza Sabouri, Haskayne School of Business, University of Calgary, 2500 University Dr. NW, SH132, Calgary, AB, T2N.1N4, Canada, alireza.sabouri@haskayne.ucalgary.ca, Tim Huh, Steven Shechter We propose a model for allocating red blood cells for transfusion to patients, which is motivated by recent evidence suggesting that transfusing older blood is associated with increased mortality rate. We study the properties of blood issuance policies that balance the trade-off between “quality” measured in average age of blood transfused and “efficiency” measured in the amount of shortage. Based on our analysis, we design efficient issuance policies and evaluate their performance. 3 - An Empirical Study of Adding Physician Assistants to Critical Care Rapid Response Teams Carri Chan, Columbia Business School, 3022 Broadway, Uris Hall, New York, NY, 10027, United States, cwchan@columbia.edu, Mor Armony, Yunchao Xu Physician assistants (PAs) can sometimes be cost-effective alternatives to physicians in healthcare systems, but their impact on critical care delivery remains unclear. Over the course of 18 months, PAs were incrementally added to the Critical Care Rapid Response team at a major urban hospital system. We empirically measure the impact of part-time versus full-time PA coverage. We find that adding PAs can reduce the average time-to-transfer for all ICU patients and reduce mortality risk for low-severity patients. Interestingly, we do not find evidence that the benefits of having PAs further improves patient outcomes when adding PAs on non-weekdays. 4 - Gate Keeping under Uncertainty: An Empirical Study of Referral Errors in the Emergency Department Michael Freeman, INSEAD, 1 Ayer Rajah Avenue, 138676, Singapore, mef35@cam.ac.uk, Susan Robinson, Stefan Scholtes We study admission decisions made by physicians in a congested emergency department (ED) using data from 600,000 visits over 7-years. As the ED becomes congested, we find that ED physicians respond by admitting to hospital more patients for short observational stays while also making fewer discharge errors. This causes a bullwhip-type effect: demand surges in the ED lead to relatively greater demand pressures in the hospital. We then show that replacing the direct one-stage (ED physician to hospital) referral system with a two-stage process in which ED physicians can stream patients with high diagnostic uncertainty to an intermediate “semi-specialist” referral unit can improve routing decisions. 332E Business Model Innovation Sponsored: Manufacturing & Service Oper Mgmt, Supply Chain Sponsored Session Chair: Simone Marinesi, Wharton, Philadelphia, PA, 75003, United States, marinesi@wharton.upenn.edu 1 - Dynamic Contests in the Lab Mohamed Mostagir, University of Michigan, 701 Tappan Ave, Ann Arbor, MI, 48109, United States, mmostagir@yahoo.com, Yan Chen, Iman Yeckehzaare We develop Knapsack Contests. An experimental framework where agents MD15 wants to reduce frequent Emergency Department visits. 2 - Issuing Policies for Hospital Blood Inventory

241